Re: Kessler effect
Kuiper at 590-630km looks like an orbital life of 30ish years. You are probably thinking Starlink (336km) which will burn up in under a year if the thrusters stop firing (and far less time when the last of the propellant is used for a de-orbit manoeuvre). You can find the numbers here.
If the whole Kuiper constellation crashes into each other then existing GPS (20,000km) and GEO (35,000km) satellites are completely safe (although replacing them would become more difficult). One Kuiper prang puts some fragments into an elliptical orbits. It is remotely possible that a fragment gets a huge apogee very near GEO or GPS. Life time depends strongly on perigee (600ish km) so this fluke fragment has under a century to be involved in another collision. That collision has to be near apogee (where almost all the other fragments aren't) and by amazing coincidence has to provide an impulse in the right direction to circularise the orbit. This high orbit is stable so there is a remote chance that with millions of opportunities this double fluke fragment will eventually hit something before the sun expands into a red giant.
Gravity was set in LEO where space is only very big rather than GEO (which is absolutely huge). Orbital period is much shorter so chance of collisions happen in years rather than millennia for GEO (or 30 minutes so Sandra Bullock does not die of old age before the next collision).
Kessler syndrome is a real danger in LEO and most nations have made an effort not to do anything really silly that would significantly increase the chances of a cascade. Elon re-negotiated his license to a lower orbit to dratically reduce the chance of a cascade. I hope Jeff does the same.