My issue was not with if the access points made sense or not... my issue was with the idea that a few grand, one time, was a lot of money to keep a wifi hotspot that wasn't on an existing network running for very long.

It wouldn't occur to me to just give people low end data plans, because for office environments? even the very high end data plans are not acceptable compared to a quality wifi setup. You might very well be right that the low end data plans are good enough for this use case, that the hit you take in reliability and accessibility is worth it to save some money.

This is actually... a problem that I (and I think a lot of technical people have) when looking at solutions; we want a good solution, and our every instinct shouts in horror when you suggest something trashy that is going to be broken half the time. But... as you point out, there are a lot of people who's expectations are so low that the low-end cellular data plan seems just fine.

It certainly seems to me like if you could get a good municipal wifi setup, that'd be worth a lot to the community. Of course, most of the 'community wifi' attempts I've seen have been failures, because making wifi work at scale with random users is a lot harder than it looks, but I think the win is big enough that it's worth trying again every few years, as we gain better wifi technologies.

But really? I'm a sysadmin, not a social worker. I can tell you a lot about setting up wifi access points, and about getting network access where there isn't any network access. I can't tell you what issues you are gonna have trying to give homeless people internet access, or what issues you might have trying to give them special phones (all the really cheap cellular providers around here don't allow you to bring your own device) or how reliable network access has to be in order for it to be useful to the homeless population- I just was offended by the implication that setting up a multi user WIFI AP for random users in places that don't have wired connections is cheap or easy.