We're talking about megaconstellations for communications, you want lower for latency, stronger signal (denser/less distance for beamforming) for better data through put -> less satellites for more coverage, and costs is cheaper since less energy. Realistically starlink has combination of 340km-1200km satellites working together, but the critical point is SpaceX reserved a lot of the sub 500km orbit slots with ITU (UN agency who manages orbits), so PRC competitors have no space real estate to try to throw up another mega constellation that can mimic spaceX economics due to location, location, location. Hence PRC registering Thousand Sails at 800km, Guowang at 500km-1200km orbits, etc, which according to OP is exponentially bad for Kessler (I have no idea). So either ITU opens much more 400km slots, or all the megaconstellations going forward going to satuate >500km LEO. Part of the reason PRC rushed to announce their megaconstellations before they even had reusable was to reserve the next closest available orbit slots that they can.