As long as there is no working ranking system every classification of a pro is just an opinion, an opinion which is equal to any other player's opinion.
The bold part isn't true. Right now we've got a ton of different players / groups of players who all think they know how to play this game well, but no way to compare to other players conception on playing well, and therefore subjectively classify themselves as pro and other players as either also pro or noob. Classification of other players is significantly based on how similar those other player's conception of playing well is to one's own. However, because of the relative lack of high-level games (read: lack of team automatch), wherein different conceptions of playing well would come into conflict, there is very rarely an opportunity to observe and compare the different conceptions. This leads the subjective opinions of pro/noob, but catch is that some of those conceptions are better than others. We just have little way of knowing which ones are which at the present time.
Despite these subjective difficulties, the OP's point remains. When I'm hosting a game, I now tend to look up each player' record unless I know them already, and if there's something wrong with it, they get kicked. Things that could get you kicked include outrageously low win%, very low total game count, etc. Is this a foolproof method? No. There is a risk I might kick a decent player. But these are calculated risks - the probability that a decent player has such a terrible record is very low, and must be balanced against the probability that the player with a terrible record is indeed terrible. The latter case is far more likely, and threatens to ruin the game for everyone if left unchecked. It would sure help if joining players would help the hosts out by realizing that their record strongly correlates to their own level of gameplay.
