The naturalistic fallacy is a fallacy of sorts, because we ascribe a deity-level "ought" in the absence of any evidence that there is such a thing. I'm not sure Wilson quite nailed the argument against it, and I'm quite sure Sam Harris, who also objects to the naturalistic fallacy, didn't nail it (in anything of his that I've seen anyway--I've seen blundering straight through it by appeal to obviousness without display of any philosophical insight).

There is a very special kind of is that bridges the gap between is and ought when you assume materialism.

That is, it is the "is" of our "ought". What is our "ought"? From whence does it come and how is it implemented? Is it goal-directed? Can we use reason to balance competing interests? Can we use the usual tricks of generalization and observation on it?

There is no ought but the ought which is--as soon as you posit something else, you've left materialism behind, and also left behind any evidence that you might use to support your point of view.

So we can (potentially) understand our oughts, and use our higher cognitive functions to extend them to be minimally contradictory and maximally achieved and so on.

If you object to this line of inquiry by saying: but why ought you do that ought, the proper reply is simply: "but what on earth can you possibly mean?" Because we cannot measure oughts except via what our oughts actually are, extended by our ability to reason.

I think Patricia Churchland may have stated a similar point of view, though if so I forget where. I do know that a variety of philosophers have tried to recast all oughts as conditional, but I don't think that actually solves the problem in the same way that the is-of-ought self-referentiality does (which is my interpretation of what Wilson was getting at).

If you object by saying: "what if our oughts turn out to be horrible? Surely this is wrong?" Then the answer is: "but this would simply reflect a failure of reason--because if it is horrible we say that it is horrible because our oughts tell us that this ought not be; and so if there is anything better, we have gone awry in our reasoning."

Now, there is a limitation to this view of oughts: you're not guaranteed to have an answer to any given question about whether something ought to be done or not. Indeed, if we don't understand the is of our oughts very well, we might conclude that we cannot answer any question of oughts because we simply don't know what we're talking about. Or there might be multiple equally good solutions. But that's just the way things go when you're constrained by evidence: sometimes the evidence isn't there, and most times you still have to retain the admission that it is possible that we are wrong.