Your poiunts are unarguable for the most part, but I have a few points:
1) I've never heard of a species becoming extinct because it was too successful. The graphs I've seen of lynx that over-hunt their prey, for example, show an increase to a peak, then a die-off,, before equilibrating around some lower value short of teh peak (or undergoing fluctuation). So I don't think human extinction is exceedingly likely.
2) Lemmings act in the interests of species survival when they commit suicide, but this is merely a lucky happenstance. it is not a conscious behavior, nor is it something that evolution selects for. There is no evidence (or very little) of any behavior having EVOLVED in the "interest of the species". (obviously CONSCIOUS behavior is excepted.)