Well, look, you're sort of coloring the debate by deciding that certain stats are "ridiculous". What makes a stat ridiculous? That you haven't used it before? I mean, for as long as the process has been around, stats like "714 home runs" or "511 wins" or "2,130 consecutive games" or "1.12 season ERA" have been a part of the discussion. Why are those stats allowable, but newer stats "ridiculous"? (I should point out that the only remotely "advanced" stat that I've been using when talking about McGwire is OPS, which you get by taking two things that've been on the back of baseball cards for decades and adding them together. This isn't exactly vector calculus here.)
But again, the point is, you think Raines should get in because he was the best leadoff guy around other than Rickey. But what if I say he wasn't? What if I say Vince Coleman was, for example? Or Lenny Dykstra? Now, I'd be wrong, of course, but how would we discuss this further using the framework you're proposing? You'd say "Tim Raines was the best leadoff hitter in the NL for years and years" and I'd say "no he wasn't" and you'd say "yes he was" and we could go back and forth and never get anywhere. If you're not allowed to tell me that Raines had a career OBP of .385 and Coleman had a career OBP of .324, and that Raines stole bases successfully 85% of the time and Coleman stole bases successfully 81% of the time, how does the conversation advance beyond simply contradicting each other?