GFA I understand your point. Basically to remove all variables before testing, which in essence is probably the best move for this experiment. I just think you would have seen a difference either way.
Still think this thread is excellent GFA! Nice work.