Ok, my final post here for a bit:
Just to clear up this definition of 'natural selection'.
I do say that all evolution can be understood with this simple concept, but the concept can be applied at different levels to different reproductive units. So, it is often applied at the level of the individual organism, but then with social insects for instance, it only makes sense when applied at the level of the colony (a 'superorganism').
However, I see nothing problematic about applying the concept of natural selection at the group level, something which some readers of Gould think that he defined as something separate from natural selection. I don't think Gould was entirely consistent on this, and perhaps he did change his mind at the end of his life. If so, I think that change of mind was a mistake.
So, it is quite possible to conceive of two rival groups competing for the same land, say. Let's say that they are human groups and that they fight a war, a war that will leave one or other group totally destroyed. One group has a 'Sparta-style system in place whereby all weak children are killed and all the surviving children are trained in a harsh militaristic, anti-intellectual way. Such a group produces fearless fighters but perhaps fewer innovative thinkers. The other group does not kill off the physically weak and places as much value on the development of the intellect as the development of the body. This group contains greater genetic diversity than the other group, finding a place both for the strong and the clever.
In the war between the two groups, whichever wins will be the reproductive unit that survives and passes on its way of doing things, which in turn will affect the future genetic makeup of the people of the area. That's still natural selection. The winning group may win through an apparently outrageous piece of luck. But again, that is still natural selection, and over the course of many such occurrences, luck will become an increasingly unimportant feature.
And such levels may be operating all at the same time. There's no contradiction or problem here. It's just a richer and more complex picture than that which is sometimes presented by those who deny the validity of one or more of the levels at which selection can take place.