We could have various different "high scores" – one for users who file the most corrections would be a simple measure. Another one for users who reject the most corrections could be funny – it isn't clear that rejecting corrections is a good way to improve your reputation.
Other categories could be worth including. (Most articles, most postings, etc., etc.)
I also just had the thought that rejected corrections that have themselves been rejected (i.e. the rejection is rejected) could assume a special status. Maybe a poll is automatically created to decide who is right. As for what effect the results of the poll would have… I dunno. Something to think about. (These ideas may be the basis of a useful alternative to the "card system".)
This also reminds me of some discussions a few months back in which we spoke of how a more meaningful user/entry rating/score might be a multidimensional object. Corrections accepted and rejected might form two components of that entity. --rspuzio 1 May 2005