Yeah, MK overcentralizing the metagame is only a smallish issue, imo.
Think of it like this: if people played Brawl in order to stir up numerous amounts of lulz, rather than go for the prize money by actually winning the match, everyone would pick Captain Falcon and he would subsequently become the center of the metagame. Doesn't necessarily mean he's a good character, though.
Of course, I'm not saying that metagame centralization is a good thing or anything; it does have its own merits in degenerating a game.
If the data for "usage" and "winnings" vary differently, aka high % of MK used and low % of MK winning OR low % of MK used and high % of MK winning, then it won't really be necessary as that data alone gives us a correct conclusion(Former means MK is really bad while latter means MK is really good).
If we see a low % of MK use and a low % of MK winning tournament money, then we can just go ahead and write it off as "MK sucks."
If we see a high % of MK usage and a high % of MK winning in addition to that, things get complicated, because then we need to go back and check MK's overall placings and find a formula to put MK's low placings against his high placings and find it out from there. Finding a formula for that would be piss annoying though because every tournament has a different entrant size so there'd have to be something in there to compensate for it.
Of course, this applies to any character, but I'm mentioning MK simply because he's the only one with skewed data at the moment. >.>