Of course the morality comes from the reality somehow. But we often don't know how exactly. But there is a pressure to provide an explanation. So people invent wrong explanations for their morality.

Later, when the official explanation is proved wrong, people are generally bad at understanding difference between "the specific explanation E of the moral norm M is wrong (but there may be some other explanation why M is good)" and "the moral norm M is bad". So the proponents of M are typically reluctant to admit the mistake in E.

Also, it's not just about truth, but also about politics. Perhaps E is wrong. But for a long time it was successfully used to defend M. Not all explanations have the power to convince people. It may be politically wise to keep a wrong, yet convincing explanation, instead of replacing it with a less convincing one, or even admitting that you don't have a good one.

EDIT: The political aspect is complicated by the fact that convincing explanations must have short inferential distances (for their target audience). The true explanation may be too difficult for this. The long-term political solution to this problem is to change education, to make the desired inferential distances shorter.

Doesn't the "is-ought fallacy" normally simply refer to the fallacious inference that because something is the case, it therefore ought to be the case? Maybe I meant the naturalistic fallacy.