Is there really a problem out there of people saying "I'd really like to play a less powerful character than the rest of the group, but gosh darn it the rules won't let me"? Is this something that comes up so often that it's an issue? Because I have to admit I've never really encountered this problem, except in a very specific sort of way.
I don't think I'm the outlier here, and with that in mind, I'm really leery of the idea of using the idea that people want to play less capable characters as the basis for discarding the entire idea of balance. It feels like the wedge point of a crowbar sort of argument, setting you up so that once you agree balance isn't important for the inevitable "So you won't mind if..." that gives the GM's girlfriend a demigod to play.
In my reality at least, balance issues come up because of one of two things (which are in some ways the same thing) - a player wants his character to be more powerful, or he wants him to be more unique.
More powerful is actually the easy one to solve, since it is usually a function of players who have spent too much time playing games where you start out as a sack of suck. Very few modern games support that model anymore, and even the worst offenders of the past (D&D and White Wolf) have moved beyond it. Most games allow for a starting character to be capable enough to satisfy most itches.
You'll occasionally bump your nose up against certain absolutist ideas like "I want to be the best swordsman in the world." Some systems support this out of the box, others do so with minimal tweaking, others are just going to jam up on it (4e, for example, simply cannot usefully support that concept). When you encounter an idea like this expressed clearly, it's usually pretty easy to work with, so long as you choose the right tool.
But all in all? Power is easy. Uniqueness is much more complicated.
Right off the bat, uniqueness can be hard to spot because it can look like a grab for power. A player looking to take powers that are not normally available, or mixing inappropriate magics or the like may seem to be trying to grab power but that's not always the case. In many cases, a player will try to pick something pretty far outside the box of playability for other reasons. Some players just gravitate towards a type, and try to shoehorn it into everything. Others have particular ideas about creativity they're expressing in this way. Others want their character to stand out in some way, and are going for uniqueness.
This is where the most subtle challenge to the very idea of game balance can really come to bear. Some of these concerns are easily addressed: players who want their character to be unique have often faced the same bad experiences as the people who have dealt with weenie characters. Again, modern games tend to do a better job of putting characters into the spotlight, so there should be less need to get attention through extreme measures.
It's the other motives that are more problematic, because they speak to player motivation, and as noted yesterday, balance is REALLY all about the players and their level of engagement. The player who wants the weird, far out thing is probably excited about it, and excitement is powerful currency, but it's how we end up with Kender.
Kender, for those unfamiliar, are the halflings of the Dragonlance setting. In older editions, they're a mechanical nightmare and more problematically are all fearless kleptomaniacs. This means that when there's one in your group, you can expect them to promptly grab the spotlight and keep it permanently affixed to themselves. Try to focus on someone else? All the Kender needs to do is steal something from a party ember or do something stupidly reckless and it's back where it belongs.
Now, it's easy to say that the kender issue is a social one, and needs to be addressed by speaking to the the player. That's certainly part of the issue, but that lets the game itself off a little bit too easy. If the game is going to leave kerosene and fireworks on the table, focusing entirely on the player's matchbook solves only half the problem.
And that's where this comes back to that elusive issue of game balance. One other reason to keep a game balanced is to keep any one player from hijacking the game (or at least to help give them less of a shield to hide behind when called out on that behavior). And that's where we get into some interesting social territory. The character with the greatest ability to hijack the game is rarely the most powerful one (unless the power discrepancy is truly huge) and may even be the weakest one if balance is purely mechanical. Note that while the kender has some mechanical issues, it's real problems are separate form that. A mechanical balancing system won't stop a behavioral issue, at least not directly.
Now, as I said yesterday, no system is going to solve all these problems, but it's worth understanding what these problems are so you can better assess whether the tools you have are going to help or hinder the process.
1 - There's a reasonable argument that 4E has not moved as far beyond it as it first appears, but characters do start with more options, and can no longer be killed by a bag of cats. I call that progress.
2 - Because, of course, forbidden combinations with obscure powers are usually where the genuine abuses of a system come up. Thankfully, you can usually tell when a player is picking combos for abusive reasons because, well, it won't be the first time. Say no or talk about it or do whatever you need. The only time it's a real worry is if your player tries to trick you, pretending they're not looking at combos. At that point, their ass may need a good bouncing.
3 - One great passive-aggressive attention-getting tactic is to play the victim. All a player needs to do is throw himself into situations where he needs to be rescued to keep the focus on himself, and he can hide behind his weakness if challenged.
4 - One more excellent argument for balance existing on the player level, not the character level.