It seems to me that given the reluctance to award errors, we should consider a player who accumulates them to be pretty unreliable. You don't get cheap errors any more, with the possible exception of an OFer who's throw hits a runner and skips away. So fielding percentage, while not a be-all, end-all measurement, has value. If they were awarding errors willy-nilly, then you could ignore it. They don't. They are clear, unambiguous indications of plays that should have been made. If we consider a SS who gets to 15 extra balls a year to be great (by UZR), then we should also consider extra errors.