You state that Randolph was not at Monticello when some of the children were conceived. I’d have to see very convincing proof to believe that (preferably from sources with no political agenda). Even if true, though, there’s no reason to assume that all her children had to be fathered by the same Jefferson. For that matter, any slave in the vicinity who was fathered by a Jefferson — in that generation or a previous one — could have passed on Jefferson DNA to a Hemings child.
I think the proper attitude on this matter is extreme skepticism. Though the modern charge had been brought up earlier by the black historian, it wasn’t until the Clinton-Lewinsky scandal that a concerted effort was made to use it to discredit Thomas Jefferson (in an attempt to show that Bill Clinton wasn’t so bad after all). In the historical article that accompanied the publication of the DNA findings in the journal Nature, Joseph Ellis explicitly made the connection with Bill Clinton. (Ellis, by the way, was the Jefferson biographer whose lying about having served in Vietnam was later exposed.).
Fawn Brodie started writing about this many years before Clinton.
To me, it’s not a Clinton thing at all, it’s history. Martin Van Buren had two black daughters. Shrug. It’s history.