Just one comment about point 4. I think 0 is only likely if you expect to create it and that customers will come on their own. If you create an SaaS that solves a real problem then you are definitely going to earn money it just won't be passive. Operations, support and marketing might become your new full time job but like acangiano said, depending on what your SaaS is targetting there's no limit on how much you can earn.
The bottom is -10k, you have to outlay for lawyer, marketing, hosting and possibly travel or industry specific capital. No matter what your research you can have zero paid uptake in the market. I.e. Google can solve your chosen task for your chosen niche for free and by accident or your domain experts can be confused about what technology will solve their problem until it really exists and is in front of them.
The majority of SaaS applications do not require their founder to travel or purchase capital. You have a good chance of not needing a lawyer[0], especially if you already have an LLC like a large number of developers do. Marketing is a likely expense but can be done for free if you're smart about it.
Hosting is really the biggest up-front expense, especially if you expect to have large spikes in traffic (SaaS around events or seasonal things) that you need to try to plan for.
[0] Meaning a large cash outlay to a person. You can get away with generic privacy, TOS, etc.
You can suceed on $0, but are you willing to fail to keep that limit? I've never seen someone not sink $5-10k post release into a production stage flop. A success may never need paid marketing, but a flop is a "what if" until you tried everything that seems reasonable.
The LLC is also not a total firewall. You can always raise your top by having the wrong/no insurance, but that is just making the worst outcome further negative.
I'm not saying its bad to try your own buisiness, I'm saying its silly to think your only downside is lost time. Unless you are a social pariah that no one will assist, you will have a way to risk losses and you should think about that and try to set limits that are realistic enough to stick to.
