[ssivark]

> This system is completely stable—I double-checked with computer simulations. But nature would have a tough time forming this system.

If I understand correctly, I think the author uses "stable" to just mean a _fixed point_. The fixed point must actually be unstable to just about every possible perturbation, which justifies the second statement; if the fixed point were actually "stable" (to perturbations) then it would occur relatively easily in nature.