My statements are based on watching Melee develop over 5-6 years and seeing things in 2007 that NO ONE predicted in 2003/2004. You are saying the past won't repeat itself. I'm saying, give things a chance.
Assuming you're right even, what good is it?
First, there's no way to predict someone rising to an effectiveness to rival MK. Even if they do, you can't say they'll still have anyone else who can counter them -- it seems most likely anything that lets them defeat MK will break any current counters they have. If they rise above everyone but MK the situation is no better than it currently is, if they rise above MK then we're likely in an even worse situation and they'll have to be discussed for banning.
Second, you haven't explained the harm caused by banning MK prematurely. We're already losing players who are tired of fighting him constantly, due to his lack of counters, lack of disadvantaged stages, and sheer number of people playing him. So we'll lose some people who only main MK, but stop losing people who are tired of fighting him
and others will return when they hear about it. Other characters will suddenly get increased playtime, and the possibility of someone unexpectedly rising to dominance will raise dramatically. Nobody will be returning that has left and the chances will continue to lower of someone unexpectedly being found to be awesome, if things stay the same.
Third, he can be unbanned if someone does rise to dominance and can be demonstrated to break his chokehold. If they can't, that new character may have to be banned as well -- but that would require quite the amazing technique application to reach that level of power. I'm not ready to accept that even as an unlikely possibility without some actual evidence of anyone having it. Not even MK really does -- it's no one technique that makes him so effective, it's the sheer number of
good options he has.
tl;dr: There's little to be gained from leaving him in, there are gains from removing him, and the removal can be undone if something is found in the future. That last one is why it's not required we be 100% sure nothing will be found, just reasonably likely is enough.