Well, the reason I said that was because no evidence was being given for causality other than correlation. So the fact that they do not always coincide was what was needed to dispute the argument given. The fact that correlation CAN be a result of causality is completely irrelevent in this case.