I have to say, I found this thread to be quite amusing. I have to assume that you are simply putting on an act to get a rise out of people. The other option would just be sad.
"This form of logic" that you refer to is fundamentally flawed. I assume you know this since, once again, I have to believe that you are joking. But just in case:
Your extrapolation has no scientific basis at all. You are performing a linear extrapolation of each single statistic from a system that is highly non-linear in nature. What exactly is your basis for this? A couple of examples (Moorman passing, etc.) have already shown the flawed nature of this approach. Do touchdowns really scale linearly with number of attempts? Do completions? You can't use extrapolation to "provide a comparable context in which the QBs can be evaluated fairly", unless the extrapolation itself is valid. To do so is simply junk science.
You have repeatedly ignored the fact that each quarterback had roughly the same number of possesions. Brady was able to do more with those chances. You argue that Trent wasn't able to because of reasons that had nothing to do with him (penaltes, etc.). So if Trent had double the number of chances (possessions) that Brady had, his performance (productivity) would have been better, making him the better quarterback. Surely you see the flaw in that argument.
You are attempting to judge quarterback performance solely based on statistics. This is no better than a quarterback rating (which someone else has pointed out). Judging based on statistics alone is pretty silly. A quarterback's stats may be reported as his, but they're obviously not his alone. They depend on everyone and everything around him. Surely a quarterback's performance has to be judged beyond statistics alone. The ability to lead, and the ability to step up one's game when it matters the most. Brady was able to do this, Trent wasn't.
Does that make Trent bad? No, he was actually quite good. Not better than Brady, though.