It's a different point, though I guess it's not utterly different from Rand (or Harris).

The point is not that the classic is-ought problem is solved. The point is that if we don't allow ourselves to postulate things without evidence, we don't even have the universal Ought that is (at least implicitly) assumed by philosophy. So it's not objectively clear that the classic is-ought problem is any better formed that questions about how many angels can dance on the head of a pin.

Now, this is a little bit of a cheat in that we haven't actually proven that universal-Ought is impossible (just like we generally don't prove that Descartes' demon is impossible). It's rather exploring what happens if we restrict ourselves to that which has some evidence.

So, yes, the self-reinforcing pattern of perpetuation is ontologically prior to the cognitive conception of "ought". (I hesitate to say that the goal is because a "goal" is also a cognitive thing, and I don't want to make claims about the relative primacy of goals and oughts--among other things, I don't think a vague intuitive notion is sufficient to allow an ordering to be determined, if there is one. Under some conceptions there might not even be a distinction between a cognitive goal and a cognitive ought.)

It's not the same as Rand's point because I don't presuppose a goal as a necessity to have an ought. Rather, I am simply noting that we humans do indeed have a sense of an ought: our ought is. But we're stuck simply noting that it exists unless we can do additional generalization, so we do our usual trick and ask about evolution, which gives us a goal-like pattern. That we picked up a goal-like pattern along the way was incidental rather than a logical necessity; it's simply an effective and compact way to help us understand what our sense of (moral) ought is, and to start exploring it using ordinary cognitive tools like pattern completion, self-reinforcement, and, yes, goal-directedness.

It is not the case, in any ordinary circumstance, that people "ought" to rape, because it is an instance of self-advancing but group-damaging freeloading behavior that overall harms social creatures' ability to keep existing. Freeloading is a huge problem for the development and maintenance of societies--possibly even the central problem. From the perspective of any ought that helps us gain the benefits of society (and it is what is possible via our society that is possibly our biggest evolutionary advantage), rape is way up there as intolerable: very much ought NOT do that. If you manage to warp the premises sufficiently much so that rape actually is necessary to continued survival of the species--let's say somehow the entire planet becomes consumed by incest-promoting cults, so that genetic diseases threaten human existence, and there is literally no more peaceful way to fix the situation--then I think it's far less clear that people "ought" not rape. The entire scenario is so fantastical and bizarre that it is hard to trust one's intuitions (or virtue- or rule-based ethics crafted under more normal circumstances)--but note that we do, generally, allow these sorts of seeming-exceptions when they come up more often (e.g. murder-in-self-defense).

The thing is, if you take seriously the idea that a sense of moral ought evolved because of its useful function, you can't read out random behaviors or goals and say, "well, that happens, so you ought to do that". You have to meet the phenomenon at the same level as which it is selected for, just like you don't see someone with a bad tremor in their hand and say, "Oh, hands are made to tremor". No, that's not what hands are for; that's a breakdown in the full normal function. (Racism and sexism are a little more complicated; I will address those if you want, but I've focused on rape because it's very clear, functionally, that it's opposed to oughtness at the level of society, and as a society, we're going to use that flavor of ought.)

Now, you could come back and say, "well, hey, but wait, the instinct to rape evolved too, presumably; what privileges the ought-instinct over the rape-instinct"?

To which I would say: nothing really, but the ought which we have is the ought which we have, and you (society-level stricture) ought not rape anyone, and if you try, we ought to stop you if we can; and this is not a confusing statement to anyone save, perhaps, some philosophers. There might be some cases in which we can successfully argue that our intuitive moral ought is actually in error relative to the role ought is selected to play (I think conservatives hit this quite often, actually), but rape certainly isn't it. Again, we have no problem thinking about this sensibly when it's about non-oughts: if there is a visual illusion because despite being evolved to deliver reliable information about how things are, there are still exploitable flaws, we have zero issue with saying, "That's an illusion. The real case is this." Likewise, we should have no hesitation, when evidence is good, to make similar pronouncements about moral-ought illusions, if we detect any.