You seem to have very strict definition of rural that doesn't map to how the FCC or Census Bureau define the term. I live in an area that is very clearly 'rural', most of it has been built on in the past 30-40 years and the average home price within a few miles of me is ~$400k. I live by engineers, doctors, architects, technologists and none of us have had anything but 10M DSL and Hughes satellite options for Internet in the past 15 years.

Yes everything you've said about the possibilities is technically true, but it hasn't happened and there is no evidence it will happen. The RDOF has awarded another provider funding to bring fiber to within a few miles of my place but not directly to me.

Meanwhile I've been on Starlink since last February. I just checked and I'm getting 110M/12M with a 28ms ping time, which is 10x what I was getting for effectively the same price. Not only that, if I want to go on the road I can pack that little dish and take that bandwidth with me just about anywhere in the US. If I was broke I could pool funds with a few neighbors and we could split the cost.

With that available now, why would anyone spend millions of dollars to drop in a 5G tower to pick up another 50-100 customers?

Now my notion of under-served rural America might be biased. I have traveled quite a bit around rural America and by far the most people in under-served communities were on the poorer side. Now I did definitely see affluent communities with minimal infrastructure (particularly in California), but that was very much the exception (even in California).

> With that available now, why would anyone spend millions of dollars to drop in a 5G tower to pick up another 50-100 customers?

Your county would—and should—spend a few million dollars with some extra funding from your state and local telecom companies to drop a few 5G towers to pick up a few thousand more customers and voter happiness for sure. Your county counsel can call it something like Rural Broadband initiative and have a plan to bring a broadband connection to 30-70% of under-served rural households in the county before 2030. A good politician should be able to sell it even to city folks as serving the community.

that's true in terms of average income, but $110/mo USD for starlink is about the same as a 20GB quota per month consumer geostationary service that is absolutely a terrible and miserable experience (viasat/wildblue/hughesnet etc). Or at most, $10-20 more per month.

> Under-served rural areas in the USA tend to be pretty poor, and as is starlink (or starlink-like) internet is pretty expensive both for the service fee and hardware requirements.

It's a utility. If it costs twice as much as elsewhere you still need to pay it, until it starts getting really difficult.