One way the metaethical naturalist might be tempted to respond to the Open Question, Triviality, Normative Knowledge, and Moral Twin Earth arguments is to appeal to Lewisian "elite" properties (which represent objective natural similarity / unity, and function as "reference magnets" when linguistic dispositions / "use" underdetermine meaning). According to what we might call Elite Moral Naturalism, 'right' refers to the most natural (/simple/"elite") natural property that systematizes the moral platitudes that capture our common use of the word. Suppose this turns out to be the property of maximizing happiness. Elite Naturalism avoids the standard anti-naturalist arguments because it's a non-analytic and yet observer-independent fact that maximizing happiness is a more "elite" property than other moral candidates (conformity to such-and-such list of duties, etc.). Nonetheless, I consider it an unappealing view. Let me explain why...
Firstly, some might question whether appeal to this mysterious higher-order property of "elitehood", over and above the qualitative natural facts, really fits with the spirit of traditional metaethical naturalism: "Haven't we just replaced one kind of magic halo with another?" I feel like there's something to this concern, but the Elitist can at least respond that we're already committed to elite properties (insofar as we want to privilege green over grue, plus over quus, etc.), so the view is at least more metaphysically parsimonious than the non-naturalist's insistence on taking normativity to be an entirely new primitive category of being.
Next, one might put pressure on the proposal by noting that in the paradigm cases of elite properties (green vs grue, etc.), the elitehood facts are fairly transparent to us. Reasonable people don't argue about whether green or grue is simpler, whereas we do argue about rival moral theories. But perhaps not all elitehood questions are as straightforward as the paradigm cases. Biologists might argue about whether the most natural definition of 'life' is one that extends to viruses, or only to more self-sufficient reproducers. It's not clear to me that there's really anything substantive at issue there (i.e., whether we can speak of "the" uniquely most natural biological classification, as opposed to various candidates that are more or less useful for different purposes), but perhaps it's enough to give the Elitist some wiggle-room.
I think their central problem is instead found in the flat-footed objection that Lewisian elite properties just aren't normative enough. Note, for example, that the Open Question Argument arises all over again: "I know that maximizing happiness is the most natural systematization of our moral platitudes (say), but is it right?" When I wonder what I should do, I'm not just wondering about taxonomy (or the comparative simplicity of various candidate classificatory codes), the way that I am when I wonder whether viruses are alive.
Compare the analogous view in philosophy of mind: the Elite Materialist holds that consciousness just is the most natural way of systematizing our ascriptions of consciousness. So when I wonder whether a silicon-chip-based "brain" would be conscious, I'm really just wondering (something like) whether a functional description of my brain is more natural than a biological description. But such a view seems clearly wrong: Consciousness is something in the world, not just a way of classifying what's already there. Similarly, I think, for normative talk. Facts about Lewisian naturalness/elitehood might be taken as evidence for taking the normative properties to track one natural property rather than another, but these facts do not straightforwardly settle our normative questions, because we are asking about something over and above the classificatory question.