I wouldn't try to code for good players in the first version.
If you get those numbers right, the good players will bring the score difference below the first if statement, and there will be no auto balance.
Perhaps the finale could have different logic applied, but I can't really see making a case for that for the first roll out either.
Like anything, see how it affects the game and adjust as needed + see what other players say.
Do you keep logs of the final game score or for each level? That would be useful for optimizing this.