First of all, I don't play Chun-li and I have no desire to play her. Of course, I have played against many, many, Chun-lis.
I don't think you would be able to find many people who don't cite Chun-li as the strongest character at the moment (maybe Herr, but along the same lines both of these characters do something that aborts the mechanics of the game and with little cost or effort, the latter obviously aborting the mechanic which is the control check and relative risk-reward mechanic the game rests on).
Has the Chun-li character 'become' to good? No. She hasn't really changed at all... She's been good since release. The fact that she wasn't errata'd/banned immediately indicates that what she is measures up to her design intentions. (At least there was one design restriction on her ability, that being that she can't play actions with her R: ).
To answer the main question by the OP, a character is 'too good' when it is seen by many as the only option (or one among a select few, all of which would therefore be 'too good') to winning a major tournament. This understanding leads to an abundance of sandbags and diversity among the top cut of players.
I guess the big question is, 'have we reached that point?' and will we even be able to answer that until after the major summer events are over? I know that your post is attempting to ask that with it's data, and I think most of us will be hard pressed to look at the data (even those that play Chunster religiously) and not say that she is markedly better than most other characters.
- dut