That's well-analyzed. My point would be that Haidt's changing of his model is consistent with his having made a political/strategic rather than a clinical decision since, as you point out, the alternative is that libertarians aren't so moral.
And it's hardly counterintuitive to suggest the latter possibility. Think of the Tea Party libertarians who shouted at Ron Paul that society should let the uninsured die if they need medical care. If the emphasis on individual liberty amounts to egoism, that's arguably the opposite of a moral outlook. Haidt obscures that fact with some special pleading on behalf of libertarians. They get their own "moral foundation" in his model because they complained.
But we could easily invent foundations out of thin air to explain away all forms of apparent immorality. Would Nazis score high on a test for Volkischness or for loyalty to their Fuhrer? How about a special moral foundation for serial killers who show great attention to detail and courage under pressure?
No, Haidt's approach seems arbitrary to me.
