At low-band, it's often quoted as 20%, but when you get to mid-band it's a lot more. That's one of the reasons there's been a mad-dash for mid-band spectrum.

I'd also note that the radio standards are made to evolve. LTE wasn't nearly as efficient as it was when it was introduced which is why, for a time, HSPA+ networks were able to compete with LTE.

> but of course the telcos are now realizing this spectrum, even the half gigahertz they stole of C band, doesn't actually propagate well around Earth instead of straight up and down

The propagation isn't really a surprise and the idea it was stolen is terribly biased. The incumbents on that spectrum were paid huge sums for it and moving expenses.

It does propagate well enough that it will be very meaningful. T-Mobile is seeing amazing gains using the 2.5GHz spectrum that they got through their merger with Sprint. They already cover around 60% of the US population with mid-band 5G. 3.5GHz won't propagate as well, but there are ways to overcome that like having the towers transmit at higher power and then having lower-frequency spectrum handle the uplink from the handset.

The three carriers didn't spend nearly $100B for something worthless and they aren't realizing its propagation as some sort of surprise.

> 2G and 3G standards have their place and keeping a 2 MHz channel open for them is not going to get in anyone's way, on the towers, or spectrum-wise

Running a 3G UMTS network requires 10MHz of spectrum. 10MHz of spectrum can be pretty meaningful. CDMA uses 2.5MHz for voice and 2.5MHz for data and maybe you're thinking of that, but it's still 5MHz to have both voice and data on a CDMA network.

Things can certainly be kept around, but it means less for other stuff.