I've written previously on the way in which I think the prospectivist view of "ought" leads to the unfortunate result that we (who should know better) have to sanction all sorts of wrongdoing in light of the fact that agents have what we intuitively take to be defective evidence (normative and non-normative). The prospectivist, you'll recall, thinks that what you ought-really to do is a function not of what's best, but what's prospectively-best (where that's a matter of maximizing expectable value).
So, someone produces an argument for the error theory and we're thinking of what to say in response. The following response shouldn't cut it, but would (I think) cut it on the prospectivist view: some subject out there has evidence according to which the actions that would maximize expectable value are actions that won't serve the agent's interests but would answer to the demands of traditional morality. And, so while there are parts of traditional morality that are wrong, it's nevertheless true to say of this individual that he ought to do what would best serve the interests of others and so has overall reason to do something other than what he has self-interested reason to do. It looks as if we can generate the sorts of reasons and obligations to act that the error-theorist thinks we cannot have out of nothing but ignorance and appearances. I think that's a defect of the view. Am I wrong about that?