Yeah. There’s a characteristic time constant that’s a function of density and of fan-out.
With ping pong balls and moustraps, each triggering launches the ball on a bouncing trajectory and the mousetrap on a different bouncing trajectory. One lands with a small and nonlocal footprint, the other with a large but much more local footprint. The ball can keep bouncing each time it triggers something else, but the trap not so much.
And all these sorts of vids the density is very very high. Space is not. For entertainment, let’s rough in the math.
Starlink is deployed in a couple different orbital radii and apparently intends to use even more shells eventually. Let’s assume it’s just one shell. Let’s further assume every orbit is perfectly circular and perfectly nominal and so the shell is effectively of zero thickness. So we’re not trying to blanket a 3D region of space, just an idealized 2D shell.
The surface area of a sphere at Starlink’s current most common orbital radius is ~3E8 square miles. But let’s lop off a third of our sphere’s area to remove the polar regions from consideration. In truth Starlink goes more poleward than that. OK, so now we have 2E8 square miles of Starlink-occupied space.
Right now they have about 3500 satellites up. But let’s be generous and give them all 12,000.
So now we’ve 12,000 targets the size of refrigerators spread around 2E8 square miles. Each fridge has an area of 17,000 square miles to roam in solitude. That’s an area a little smaller than West Virginia.
You’re going to have to launch an awful lot of gravel to blanket an area the size of West Virginia in sufficient density to snag the one and only passing refrigerator. And even if that ~500 lb fridge disintegrates neatly into 8000 1-ounce fragments (or even 80,000 0.1 ounce fragments), that’s not gonna add much gravel to the amount necessary to cover West Virginia.
I don’t see Starlink being Kessler Cascadable in a timeline less than multiple years, and probably low multiple decades.