The keyword is likely. I agree that there is some correlation, but we can’t know for sure how strong the connection is unless we are given the numbers, and their lack is the reason for this math in the first place. If we assume that all straight people are cis (which I doubt), then we need not do any math - the number of straight cis people is the number of straight people. If we assume no correlation at all (which I also doubt), then we get a more reasonable number. If we assume some correlation, then we just get a similar number, but the math gets a lot messier.