Re: This is where we are now
"How would LEO beat microwave links at ground level? "
Microwave links need repeaters every 40-60 miles.
Satellite is ~200 miles up and then repeater spacing 500-2000 miles (variable) using precision laser pointing - this is covered in a bunch of patents and FCC filings
Speed of light in atmosphere: ~0.99C
Speed of light in LEO: ~1.0C
Speed of light in water/glass: ~0.7-0.6C
Speed of light in coax/transmission line: ~0.75C
Speed of light in waveguide/fibre: ~0.6-0.55C (it gets a bit complex but propagation front speed is slowed down a LOT due to the longer overall paths taken by bouncing off the sides - at the same time the phase front can appear to be travelling faster than light but it's an illusion)
Depending on how the signal is regenerated, each hop is going to add a few microseconds delay in the optical/radio -> electrical -> optical/radio transitions (and a few more if actual retiming is done).
Optical boost pumping in fibre isn't done this way and effectively doesn't have any time delays (the less electronics at the bottom of the sea the better) but the speed penalty of the medium is already a big penalty.