Does that not assume equal probability of finding bugs? Seems to me that both teams would find the same easily identified bugs (say '=' instead of '==' or revealed by compiler warnings).