Taxes are how the government funds those sorts of things, and taxes are collected from everybody. So the subsidization still happens.
Take tobacco taxes, for example. We put special taxes on tobacco because we don't want people to smoke, and we want to discourage people from smoking. Taxing tobacco effectively raises prices, which reduces demand. Various governments employ similar taxes to discourage things like soda, sugary foods, etc. This is a well-understood concept: you put extra taxes on things you want to discourage (relative to the rest of the economy).
Paying for rural telecom by putting a special tax on telecom service achieves exactly the same effect. It increases the price of service, reducing demand and reducing investment (relative to the rest of the economy, which does not bear that tax). But in theory, telecom infrastructure is something we want to encourage, not discourage.
It's even worse when you impose a general public obligation on individual service providers. For example, instead of paying for universal healthcare with general tax dollars, you could simply direct hospitals to treat poor people for free. Nobody does that--they pay for the broad public benefit with equally broad public taxes. But we do it for telecom--we tell companies that in order to be allowed to offer any service in a city, they must shoulder the burden of building out infrastructure even to neighborhoods where few people could afford to subscribe (i.e., neighborhoods that can only be served at a loss). That essentially makes it impossible to have a "minimally viable ISP." You can't compete with an incumbent by stealing away customers neighborhood-by-neighborhood. If the government simply let companies build infrastructure where it was profitable, and built subsidized infrastructure itself using general tax dollars where it was not, you could have that sort of competition.
Actually, I believe that's exactly what happens in the US with emergency departments. Although it supports your point about such a practice being distortionary, it also draws attention to the fact of the sheer complexity of all this.
That's mostly what bothers me about seemingly-simple statement or analogies about economics (or, really, economic theorizing in general), that the reality is far more complex and interconnected.
Yes. That's why we have no interstate highways to move goods around, spurring commerce.
Yes, taxes to pay for roads do indeed discourage economic activity.
This is a cost.
And the benefit that we get is the roads.
And in this case the benefits outweigh the costs.
But just because the benefits outweigh the cost doesn't mean that the cost doesn't exist.
Yes roads have benefits. They also have costs.
The benefits are greater than the costs, but we should be aware that the costs still exist.
And the benefits outweigh the cost.
In other words, if Option A is "get zero dollars", and Option B is "spend $1 to get $2", it is unfair to complain about B's "cost" (spending $1) in isolation.
So, what it does is constantly relay wires, build out to rural areas, etc. and thereby extend their network and keep it up-to-date (wires only last 30 years anyway). This I think is what we need to companies such as Comcast.
Your post actually highlights the problem. What makes 10% the proper profit margin? BT OpenReach, the U.K.'s regulated infrastructure monopoly, has a profit margin of double that. That number becomes a political football, and the political result probably isn't what most people on HN would want. People are happy with 25 mbps DSL; they're not going to vote to raise Internet rates to drive returns high enough to incentivize investment in replacing everything with fiber. That's exactly what you see in other rate-regulated utilities. People don't vote to replace lead pipes that poison kids, because they would rather have cheaper water rates; they don't vote to replace sewers that leak raw sewage into rivers when it rains, because they'd rather have cheaper sewage fees.
I'm not sure how not having cheap fibre to ge doorstep is an externality.