I don't think 5G will be restricted to cities at all. There are some carriers who are already providing pretty good service to rural areas using sub6 fixed wireless. The N77 auction (3.7 - 4.2Ghz) I think will open up some very good competitive rural options. mm-wave will probably be too expensive in most cases but not necessarily always. The reason is because mm-wave being so fiber intensive. There is often a misconception that fiber is more expensive to deploy in rural areas but often the opposite is the case. Rural areas tend to have fewer regulations and restrictions that make deploying fiber a lot less expensive on a per mile basis that in cities. It's just that cities have more potential customers so it is the obvious first place to deploy. Although it will take longer to get there, I believe that fiber will work its way to more rural areas over time.

> I don't think 5G will be restricted to cities at all. There are some carriers who are already providing pretty good service to rural areas using sub6 fixed wireless.

That's a true statement but most uses of "5G" are talking about mm-wave. There's no question that lower frequencies, already in use for 4G or adjacent to those, will get the evolutionary bump to 5G.

And don't forget to consider the need/benefit side. If your 2-7GHz towers won't be saturated, even a free pile of mm-wave antennas isn't worth the maintenance costs.

The person you're responding to wasn't saying that the majority of the US is not directly in a city, but that 5G won't be useful for the majority of that subset

A third of the population isn't something to disregard. Their suggestion is that a holistic approach could be taken. But maybe this isn't too much an issue of resource allocation (ie 5G deployment may not be zero sum with fiber deployment, particularly if the considered regions are disjoint)

Oh well, hopefully Starlink or something can help raise the floor on rural internet access quality

It's fortunate that those who live in urban areas never leave those urban areas.