In order to make a good 'balance change', you should try to identify what the source of an imbalance is. If you fail to identify it, it's a good idea to revert the changes you made in your attempt to solve things, then try again.
If you continuously layer on 'fixes' that don't fix anything, you create a mess. You rapidly lose sight of your original goal, which goes unreached.
My example, of course, is the lerk!
Its speed was decreased, ostensibly because it was too dominant, and the root cause appeared to be "the difficulty of killing it." When this didn't do anything (because in the current state of the game, people can't kill slow lerks either), bile bomb was weakened tremendously.
Now, what are we left with? A lerk that doesn't have an overpowered bile bomb (good), but that can't fly effectively (bad). The result? The lerk is still no good in the early game (before bile), and it's no good in the mid-to-late game (because bile isn't an automatic win button any more). Playing the class simply doesn't feel fun, and I can't be effective because the most fundamental, skill-indexed mechanic of the class was harmed in a knee-jerk reaction.
Spikes and flight are the only things about the lerk that have any potential for the long-term at the moment. It's really pathetic how weak and annoying they both are from the point of view of lerk players. Even after reducing the effectiveness of spores and bile (the fisher price crutch mechanics), that's still all the class is really good for.
The path to rehabilitating the lerk is clear - start by returning its speed, and realize that it will be possible to kill it when FPS is raised, input lag is reduced and so on.
Paid for by the committee for never-ending lerk domination