While I agree with the main substance of your argument here, I have to take exception to your quoting an almost certainly inflated figure for the number of children not the genetic offspring of their mothers' partner-of-record. What you're looking at there is the percentage of paternity tests that come up negative. Paternity tests are generally conducted on children whose paternity there is already some reason to suspect. When you look at the results of studies on the populace in general, the figure ranges anywhere from 1% to 4%--which while still significant pains a rather different picture. There's a thread on the subject over at the snopes board--which naturally, I can't find right now.