I don't think this is going to solve bandwidth problems for heavily-populated areas.
- First, areas such as NYC and SF have more OTA broadcast TV stations, so there are fewer frequency bands available for WiFi.
- Second, the number of users of those available frequency bands will be much higher (millions, instead of 10s or 100s of 1000s).
- Third, the frequency bands themselves just aren't that large. VHF Channels 2-6 occupy slots between 54 MHz and 88 MHz, with a segment of that used for RC toy signals*. So figure 6 MHz per VHF channel. UHF Channels 14-20 occupy 470-512 MHz*, so 6 MHz per UHF channel also. There are only channels up to 51 available in the US; all of the other channel bands have been auctioned off for other applications*. Ultimately, there are a maximum of 49 (2 to 51, excluding channel 37) 6 MHz channels available assuming zero broadcast stations in a given area.
How many bits can you actually send through a 6 MHz channel? If you send more than 6 mbps, you'll have more bits than you have waves in your carrier frequency. Somehow I don't think that's gonna work very well (though I welcome counterarguments from people who know more about radio waves). So imagine 50 teenage girls in Chicago with iPhones trying to simultaneously download the latest Justin Bieber video at 720p to their iTVs. All of the new bandwidth is used, and there are still, what, like 100,000 other teenage girls still trying to download the same video? OK, so maybe not ALL of them would ALL try to download at the same time, but you see my point. There's only so much information that can travel by radio waves in a given frequency band in a given physical space at a given time. If we want to send more information, we have to either increase the frequency band or segment the physical space. This solution is great for rural areas, but isn't going to do anything to help 80% of the US population*.
* Source: wikipedia ("VHF","UHF","Rural")