PRN (Pseudorandom noise) in the context of GPS is just a coding standard - it's just CDMA (aka Spread Spectrum) and it allows all satellite to use the same frequency. A side benefit is that the signals can be below the noise floor, and when you apply the gain from decoding, it rises the signal above the noise floor (exactly like how you can pick out a voice in a crowded bar if you know what that voice sounds like).

GPS uses PN codes for the timing difference measurement, as well as allowing multiple satellites on a single channel. I believe the dithering (selective availability) was turned off years ago, thus the L2 channel ads only ionospheric correction (which can also be accomplished with local sources).

The FCC symbol rate limitation needs to go. It’s a hindrance on HAM radio. Just regulate it by bandwidth, or better yet EIRP PSD, but that would be tough to control.

Yes, Selective Availability was turned off in 2000 and will never be re-enabled. In fact, the latest generation of GPS satellites do not even support Selective Availability: https://www.gps.gov/systems/gps/modernization/sa/