I find it a bit counterintuitive. After all, the natural timescale for planetary dynamics should be the orbital period --- i.e., a few years. So one might thi

I find it a bit counterintuitive. After all, the natural timescale for planetary dynamics should be the orbital period --- i.e., a few years. So one might think that if the system is unstable, it will dissociate within a few years. If it doesn't do that, then it should be stable for all time. But instead the next system takes 10 years to fall apart. And the next one takes 100 years, and then 1000 years, and so on. And eventually these systems (whose natural timescales are still years) take billions of years to fall apart. You would think that if it is unstable, then it should reorganize itself into a system that is stable for all time. It's a little bit odd that every time the system reorganizes itself, it is only a little more stable than it was when it started. Somehow, planetary systems are always just sort of end up teetering on the edge of stability.