In short, there are universal morally relevant human preferences created by evolution (i.e. hunger, sex drive). That doesn't show that evolutionarily created preferences resolve all moral dilemmas.

The scientific method has hardly resolved all scientific dilemmas. So if there are real things in science, 'resolving all dilemmas' is not a requirement for scientific realism, so it would seem it shouldn't be a requirement for moral realism.

"Descriptive" statements about morality (e.g. 'some, but not all, people think incest is wrong') is objective. The only real question is whether "normative" ethics can be objective. 'people think incest is wrong' is a descriptive statement. 'incest is wrong' is a normative statement. The moral realism question is really whether any normative statement can be objectively true. The intuition pump for thinking "maybe yes" comes not from incest statements, but rather I think from statements like "humans shouldn't pick an 8 year old at random and chop off his limbs with a chainsaw just to see what that looks like." Incest statements are like pumping your intuition about scientific realism by considering statements like "Wave Function Collapse is how we get probabilistic results in real experiments." If you are wondering whether there is ANY objective truth, start with the obvious ones like "the sun will rise tomorrow" and "hacking the arms of reasonably chosen children to see what that looks like is wrong."