Even if you ignore the extra walking required, and even if public transit actually gets you close enough to your destination, public transit at least doubles the travel time required. Potentially more if your drive doesn't need to deal with traffic, which it often doesn't.
Random example: in the Portland area, a city known for both public transit and pedestrian/bike friendliness, I looked up routes from my current workplace to the university building I took most of my classes in, via either a car or public transit. Despite having a train line that handles most of the distance, that trip would take at least 90-100 minutes, depending on time of day. That assumes you start your trip at a time that doesn't require waiting for the first pickup. The same trip by car would take 25-40 minutes, depending on traffic. And with a reasonably fuel-efficient vehicle, the trip would cost less in gas than the cost of the transit fare. On top of all that, you can leave when you want to leave, not worry about which bus or train you need to catch, and not worry about getting stranded when the buses and trains shut down for the night.
This isn't an issue of worldview; by any almost objective measure, a car beats public transit, here in one of the most transit-friendly cities in the country.
(In the interests of a fair comparison: public transit does work better if you don't have the option of using a car, such as if you can't afford one, or if you have some legal reason you can't drive.)
Adding more public transit vehicles or routes wouldn't solve this problem, because the current set of vehicles aren't actually at capacity; by measures of throughput, they carry all the passengers that want to ride. However, by measures of latency and similar metric, public transit utterly fails.
Perhaps if you somehow convinced the entire population to use public transit, then you could put many times the number of public transit vehicles on the road without wastefulness, and partially solve the latency problem. But you'd have a lot more luck convincing people to use a fleet of self-driving electric cars instead.
I would welcome a similar latency-based comparison in a more transit-friendly area, as long as the comparison shows transit infrastructure getting better to compare well with cars, rather than driving infrastructure getting worse while transit stays the same.
I lived in Portland, and even though the public transit there is supposed to be good, it was basically unusable. As a European I try to take public transit whenever I can, but in this case it took me 20 minutes to work by car and 75 minutes by train. The reason was not missing train lines or anything, it's just that the train is super slow. The top speed is low, it accelerates slowly, it opens the doors for what seems like 5 minutes at every station.
By comparison, where I live now public transit is as fast or faster than driving, unless you're going somewhere not covered by the subway or trains. It's fast paced, because a majority of people actually depend on it to get anywhere. If I have to wait more than 3 minutes for the train, I'm annoyed. I'm not saying this would work in most American cities, because they're built for cars and you can't just undo that, but in European cities, yes, public transit is often better than driving. In Portland, the trains feel like they were built for old and poor people who cannot drive, and have all the time in the world to get wherever they're going.
As you mention, in the US, public transit is often unusable. It is often unusable in the US mostly because the cities have been designed around cars, urban sprawl, single-usage zoning codes and low density.
So far, from what I've seen, the cities where public transit works are cities with relatively high population density and mixed-usage zoning (ie. shops on the ground floor and housing/office space above) rather than the generic American suburbia with single-family houses. As the former model is how most large European cities are built, Europe has significantly better public transit than the US does. Cities that are built that way in the US also have decent public transit (eg: the core of SF, NYC and Boston are well served by public transit, whereas LA isn't) because cars don't scale past a certain density.
That does get into the worldview problem, yes; I have no interest in sharing a wall, ceiling, or floor, so any transportation option that only works when everyone does so does not appeal at all. That might change if apartments started coming with far more soundproofing than they currently do.
However, I wouldn't say that US major cities have good public transit; rather, they have awful automobile infrastructure such that transit looks good by comparison. And as you said, "the core of" those cities have transit options, but that doesn't mean the whole city does. I've certainly encountered much better public transit when traveling in Europe, though not sufficiently so that I'd be inclined to use it in preference.
In any case, that then leads back to the point that started this discussion: given a set of existing cities that don't have (or want) the kind of layout and design you mentioned, how do you build a transportation infrastructure? Public transit doesn't fit the current layout very well, and you can't redesign those cities from scratch even if people would prefer the new design. Hence, the many different visions of self-driving cars, which adapt more readily to various ways of laying out a city, and which work in urban, suburban, and rural settings.
It's a tradeoff. If you want to live in a low density area and have to drive everywhere, then no, you're not likely to have any mass transit other than a suburb commuter train to go downtown. The alternative is living in smaller, high density housing where there are two grocery stores, three coffee shops, a half dozen restaurants, a pharmacy, a metro station and so on within a five minute walking radius, and have a fifteen minute commute to work by foot or five minutes by bike, so there's not that much transit required.
The funny thing is, if we go back to the point that started this discussion, the city of Montréal actually has a decent public transit system, a bike share system, 400 miles of bike paths and is one of the most bike-friendly cities in North America (depending on what source you believe).
For those US cities that have low density sprawl (such as LA), there's no real way to have a transportation infrastructure that makes sense; people who are from LA will say it's fine, you just drive wherever you want, whereas Europeans will say it's not a city that you can live in because you spend your life in a car and can't go out and grab a coffee or pastry with a short walk.
Self driving cars don't really solve that problem, because you'll still have thousands of people going to work at roughly the same time, and even though self-driving cars can have higher density and traffic fluidity than human-controlled ones, they still pale in comparison to mass transit systems.
For example, look at Tokyo; it has at least five train stations that have a ridership over one million passengers per day and has a ridership of over forty million (granted, that number counts people switching transit systems multiple times). If you assume that the minimum surface area that a self-driving car could have is nine square feet (3x3), slightly more than a chair moving around would have, and that you need 5 million of them to have an adequate amount of service, just the surface area of all of these cars, bumper to bumper, is 1.6 square miles. Assuming you put them back to back, eight cars wide, that's 355.1 miles long worth of cars, which is about 40% of the length of freeways and highways in LA county.
> eg: the core of SF, NYC and Boston are well served by public transit
SF is, I think, a glaring opposite example: a city whose density should support transit, but which does not have it. Not a single part of SF is served by what I'd consider "modern" rapid transit, meaning average higher than 30 mph, frequent headways during the day (~3-5 minutes), and 24/7 service with at least some reasonable frequency (~20-30 minute headways off-peak). One smallish corridor of SF is served by what I'd consider 2nd-tier rapid transit, the BART corridor, which isn't 24/7, and has relatively infrequent trains even during the day, but at least averages an OK speed when it does run. Most of the city instead has 1920s-level transit. For example, to get from the Outer Sunset to the Caltrain station is about 45 minutes on a streetcar. This is a distance of 6 miles, which should take about 15 minutes on a metro.
No you couldn't. Walking for 45 minutes you'll go maybe half that. And in a congested city you may well not be able to drive any faster than that; part of the reason the streetcars are slow in SF is because they get stuck behind car traffic.
> Specific studies have found pedestrian walking speeds ranging from 4.51 kilometres per hour (2.80 mph) to 4.75 kilometres per hour (2.95 mph) for older individuals and from 5.32 kilometres per hour (3.31 mph) to 5.43 kilometres per hour (3.37 mph) for younger individuals; a brisk walking speed can be around 6.5 kilometres per hour (4.0 mph). Champion racewalkers can average more than 14 kilometres per hour (8.7 mph) over a distance of 20 kilometres (12 mi). ...
> Power walking or speed walking is the act of walking with a speed at the upper end of the natural range for walking gait, typically 7 to 9 km/h (4.5 to 5.5 mph).
> Many people tend to walk at about 1.4 m/s (5.0 km/h; 3.1 mph). Although many people are capable of walking at speeds upwards of 2.5 m/s (9.0 km/h; 5.6 mph), especially for short distances, they typically choose not to. Individuals find slower or faster speeds uncomfortable.
These all indicate that walking 6 miles in an hour is exceptional.
It absolutely is unreasonable; 6mph is unambiguously a jog and not many people could keep it up for 45 minutes. An average walking speed is 3-4 mph - half what you'd need to do 6 miles in 45 minutes.
This commute is actually very car friendly because most of the route is done on a highway; most people live closer so they don't benefit from that, it's just that the train simply travels faster than cars do inside the city. There's a reason most of the population commutes using bicycles and public transport rather than cars.
Your worldview seems to be tainted by the belief that public transports have to be slower than cars for some reason, despite them having dedicated paths (metro and train tracks). You just live in a place with bad public transportation.
Aha, perhaps that is the issue: if Portland, billed as mass-transit friendly, sucks this much w/r/t MT, it's understandable that people would try it and say "well if this is the best mass transit can do, then no thanks."
As a resident of Portland, my own anecdata is that transit is pretty good here. 30min from downtown to the airport for instance. As I noted in another comment, it's difficult for me to imagine a 100min train commute here. Taking the longest line (Blue) end-to-end, through downtown takes ~100min [1]
Copenhagen I think would meet the criteria of your last sentence: over the past 20 years, car infrastructure has remained somewhere between constant and modestly improved (depending on where you are), while public transit and bicycle infrastructure have both improved significantly. The new metro system (completed 2007) is an especially big improvement: it runs 24/7, is driverless, and the wait between trains is only 3-6 minutes during most times. Another nice addition is the completion/expansion of the "A-bus" network of high-frequency (~6-9 minute headway), 24/7 buses, which fill in gaps between the rail service. And a lot of new bicycle infrastructure has been created, too.
In the core urban area (central Copenhagen, Frederiksberg, parts of Amager, etc.), driving is now probably most often either 2nd or 3rd place in latency. Usually either biking or public transit will be 1st, depending on where precisely you're going from and to.
> public transit at least doubles the travel time required
That might be true in Portland (and, arguably, every American city besides maybe New York), but it's definitely not true everywhere. Go to Hong Kong and you'll see how fast public transportation can be at getting you exactly where you want to be in a fast and convenient way.
Random example: in the Portland area, a city known for both public transit and pedestrian/bike friendliness, I looked up routes from my current workplace to the university building I took most of my classes in, via either a car or public transit. Despite having a train line that handles most of the distance, that trip would take at least 90-100 minutes, depending on time of day. That assumes you start your trip at a time that doesn't require waiting for the first pickup. The same trip by car would take 25-40 minutes, depending on traffic. And with a reasonably fuel-efficient vehicle, the trip would cost less in gas than the cost of the transit fare. On top of all that, you can leave when you want to leave, not worry about which bus or train you need to catch, and not worry about getting stranded when the buses and trains shut down for the night.
This isn't an issue of worldview; by any almost objective measure, a car beats public transit, here in one of the most transit-friendly cities in the country.
(In the interests of a fair comparison: public transit does work better if you don't have the option of using a car, such as if you can't afford one, or if you have some legal reason you can't drive.)
Adding more public transit vehicles or routes wouldn't solve this problem, because the current set of vehicles aren't actually at capacity; by measures of throughput, they carry all the passengers that want to ride. However, by measures of latency and similar metric, public transit utterly fails.
Perhaps if you somehow convinced the entire population to use public transit, then you could put many times the number of public transit vehicles on the road without wastefulness, and partially solve the latency problem. But you'd have a lot more luck convincing people to use a fleet of self-driving electric cars instead.
I would welcome a similar latency-based comparison in a more transit-friendly area, as long as the comparison shows transit infrastructure getting better to compare well with cars, rather than driving infrastructure getting worse while transit stays the same.