Here's the thing. You need to take in the position of players when considering this. A weak champion, is only going to be played well by players that are just generally good at the game, this having very little effect on the champion's actual power level. Someone who can perfectly CS, keep map awareness, roam, etc is going to be winning more than someone who doesn't.
At that point, you're measuring the level of a player's skill, not a champion's strength. Mains of champions, or people good at the game, usually have +55% winrates, due to the basic fact that they're well experienced with the champion, and know how to play them effectively. But we don't normally balance by that, unless it's way overboard.
So, if you make a champion strong, you can get a feel on how every player can experience and play this champion, and from there, get down to figuring out whether he's fine or needs changes. The new champion isn't simply winning because the player has skills in game's mechanics, but because they have skill in the champion itself.