The "Kessler syndrome" worst case scenario - I'm just recalling stuff from Scott Manley videos here - I think is a 'really bad decade' where a cascade of collisions makes launching in to LEO impossible until everything settles down in 10ish years. Bad, yeah, terrible even, but possibly worth it in some sense? I mean, it makes about as much sense to me as growing subsidized corn for ethanol gas, I suppose. I'm sure someone is making money.
Space junk is not in the top 5 reasons, and with better tracking and using low altitudes (like Starlink does, where debris deorbits very quickly), we can probably fit a factor of 100 times more satellites in LEO safely.
(People underestimate how much of a difference better tracking makes to this issue. Space is huge, and the real isn’t actual conjunctions, but the fact that the uncertainty range for objects is so large that you end up getting hundreds of times more false alarms and probably-unneeded avoidance maneuvers than you’d have if you had much better tracking.)
[1] https://blogs.esa.int/cleanspace/2022/08/11/on-the-atmospher...
Of course speed and debris is a factor, but fact is - space is huge and we don't have that many satellites.