In his book Sentimental Rules, Shaun Nichols reports some fascinating research that suggests young children are hardline moral objectivists, who think that if something is morally wrong it is "wrong for real", and that this is a kind of "default setting" in the metaethics of common sense. If that suggestion is accepted, even some moderate theories of moral realism should actually be held as error theories, or as containing an element of moral scepticism, in so far as they maintain that moral truths are grounded in human nature and would not be true for alien beings with a different psychological makeup. This moderate kind of moral realism, for which I have a degree of sympathy, must hold hardline moral objectivism to be an error: if commonsense morality endorses hardline moral objectivism, then the moderately realist theories must hold commonsense morality to be mistaken.
This suggests to me that any moral theory that remains plausible on considered reflection is going to be, to some extent, an error theory. But why should that be surprising? The whole history of the advance of human knowledge shows common sense gradually being revealed as a poor indicator of the way the universe really works. It provides models that are pragmatically useful in the evolutionary and historical circumstances that shaped it - someone following these models in the past would have kept out of trouble. However, there is no reason to believe that commonsense models accurately track the larger reality of which our lives are a very small part. In fact, we have pretty good reason by now to assume that common sense is likely to be wrong whenever it strays beyond giving us the most pragmatic, everyday advice.