I agree that the Jewish family names go back for a long time, yet there is no way to tell how that family came by that name. One can't assume that because generations of people in a family carry the name Levy, that the originator was Jewish or that the name wasn't adopted at some point in time. I stand by my analysis. There is no way to prove an unbroken line back to one of the original twelve tribes.
I understand what the man who founded FamilyTreeDNA is trying to prove, and I find his conclusions to be flawed because of the intermixing of races of people going back to the beginning of time.
Most archeologists agree that the Israeli and Arab archeology and historical science is often flawed because of the eagerness to prove their holy books and their understanding of history. I see this as another example of that flaw. For those that don't know, Israel, Egypt et al, rarely let anyone outside of their communities participate in their science or excavations without first agreeing to the agenda to prove the existence of their history in the manner that they tell it. Most examinations of Middle Eastern archeology are necessarily done from an office somewhere rather than actual boots on the ground because of those provisos. If you are fortunate enough to get a permit to examine a site in the Middle East, its with an escort from the Govt to make sure that your discoveries uphold their story. Going into a scientific examination with the express purpose of proving a bias is not good science. One has to go into it with an open mind and then draw your conclusions.
I see this as another example of this man going into the science with an express purpose of proving an historical line back to the 12 tribes. I find that to be a flaw in the science since it allows for no other circumstances and there is no way to account for the intermixing of peoples and it assumes that any semitic race traced back by someone with a last name that is Jewish means that this line must be descended from one of the 12 tribes. There is absolutely no way to prove that descent without a lot of unfounded assumptions.
So, guess we're at a stalemate on this one, but I thank you for an interesting conversation!