Population density of Germany is 229 persons per km/sq. In the US it's 32. It's a lot more expensive to run the infrastructure for a country 27 times the geographic area and 1/7th the density.
Using national population density to explain the extraordinary relative suckage of US broadband is a straw man argument.
Because nobody wants regional service? There used to be regional service that was cheaper--I remember my parents signed up for it when they first got cellphones, around 2003--but it lost out in the marketplace, because people would cross a state line and be shocked by the roaming fees.
Edit: http://www.metropcs.com/coverage/ Their coverage maps indicate real coverage of a few metro areas, and roaming onto national networks.
It's just that regional carriers aren't a whole hell of a lot cheaper than the big national guys. Who wants a 10% discount for 1/100th the coverage area ?
Honestly, do you have any ideas about what causes these high prices other than density? The competition among the providers is pretty healthy, as far as I can tell.
Lower population density actually helps wireless internet -- the cost is mainly not about the transmission equipment, it's about the very limited amount of total over-air bandwidth.
Not quite true. Wireless internet is most cost effective when population density is at an extreme on the scale. If the density is low then you can use fewer towers that have larger cells and if the density is high then you need more capital for the higher-capacity infrastructure, but your backhauls are shorter and you have a larger userbase to recoup the investment. I would bet that for wireless operators in Finland the Helsinki area, Turku, and low-population areas make money while Kuopio is probably a break-even region at best...