So it is saying we haven't been able to isolate the extent of the social impact. But this applies to so many things. Do we have research that shows that most people are heterosexual that is consistent across culture and not confounded by social factors? No, we don't. There have historically been cultures where the view on sexuality, what constitutes sex, and how sex is viewed are so different than we cannot say that our current split of people's orientation is based only on biology.
But, this sort of disclaimer is not being equally applied because it would apply all psychology and sociology research. We have a few examples of research that shows just how extreme the social impact can be, but we don't understand impact enough to account for it.
Is it good science if the standards research are held to are not consistent depending upon the topic?