Define "usable frequencies". You are making pretty ridiculous sounding claims btw. I would rather believe that we can max out cell tower wired connection bandwidth, than completely saturate radio frequencies. Obvious solution, as non-professional, is to use more towers/higher frequencies. Give more information, if you want people to believe you. Go technical and say how things really are, no need for that "layman" stuff.
Modulation is already taken into account when talking about channel capacity. If you don't have any kind of modulation you essentially just have a sine wave and can't transmit any data.
The number of bits per hertz is bounded by the amount of noise in the signal. There are natural sources of noise, so a noise free channel is impossible. The mathematical basis for this is the Shannon Hartley theorem if you're up for some reading.
Higher frequencies don't provide more bandwidth. It's the width of the channel that matters. A 20mhz channel provides the same bandwidth regardless of whether its centered at 100mhz or 900mhz. The main difference is the 900mhz signal won't travel through solid objects as well.
People sometimes turn to higher frequencies for more bandwifth simply because there's more spectrum available because it's far less useful. For instance, cell towers sometimes communicate with other cell towers using frequencies around 60ghz. The downside is you need direct line of sight on frequencies that high, so it works for tower to tower, but not tower to phone.
Regardless, the first of his two points still applies:
> More cell towers