Although SMK has mostly answered you, I feel I need to give my own explanation for this.
As you've mentioned, we currently have two primary groups of characters... those who are generally recognized as being in the upper half of the cast and those who are not. However, what makes these characters defined this way is only really based on a "feeling." Based on how easy they are to use for the general public, and how effective they are at higher levels, and just the general traits about the character that people recognize as being "good" or "bad." However, none of this takes into account specific matchups. While we may be able to say "this character is good because x, y, and z," we cannot say how much better they are than any other given character. This is the truly difficult part. Without extensive matchup data you can only do so much. You can only tweak a char until they "feel" right.
To an onlooker, it seems we have accepted two different standards of balance, one for "good" characters and one for "ok" characters. To a certain extent, this is true. We know what chars are good. We know what makes them good. But we don't know how good those attributes make them against random character #27. So what we're doing is trying to keep everyone "reasonably close" and buff chars to the point where they can be considered "viable," until we can obtain accurate data.
We certainly have the ability to nerf what makes a char "good." The problem is, in doing this, we could very easily end up making the char "bad" and thus in need of help. It's impossible to know exactly how much of an effect any given change will have on a char without extensive matchup data, and what is useful in what matchups, and how it is useful. It's very easy to give a random character that "feels" they are lacking a buff to one of their moves that gives them a new tool to compete with. However, what if this new tool does absolutely nothing to help the character in their problem matchups, and instead only shifts their advantageous matchups more in their favour? That isn't helping balance. It's hurting it. That's why we can't "just" nerf the good chars or "just" buff the weak chars. It's a lot more complicated than that.
Right now, the game "feels" relatively balanced. But that doesn't really mean anything. Everything we've been doing so far has been trying to get chars relatively close to each other in terms of power, and so far, it seems to be working. What makes it especially difficult is that people have different views of how certain chars should behave and how they should be buffed and/or nerfed. When it comes down to it, it's probably more accurate to say that the WBR is only currently trying to "streamline" characters right now, rather than "balance" them. Yeah, they all seem to play "well," but until we start having more tournaments with a consistent build and people get more matchup experience, we won't know how balanced the game really is.
I'm not sure what the other WBR members think about this, as we haven't discussed it in a while, but my idea for a release schedule of brawl+ goes something like this:
For the next couple months, the WBR will continue trying to "streamline" characters, until we feel we have an extremely solid cast worthy of a Release Candidate set. For the next 3+ months following, a number of very minimal changes will be made to this RC set each month, as specific balance issues get brought up. When these one month trial periods stop yielding new tweaks to be made, a Gold set will be released, which at first will probably be updated every three months or so, with updates gradually being made further and further apart to allow the metagame to stabilize and grow on its own, without constant changes. Eventually, small updates will be made no more often than once a year, if even. The balance tweaks made after Gold will be done to change the characters in as small of a way as possible, while specifically helping in matchups they need help in, and hurting in matchups they do too well in, without helping or hurting too much in places where they don't need to be helped or hurt in. In the event that something gamebreaking is found, depending on how extreme it is, an emergency patch may be released early to fix it, but don't count on "emergency" updates getting made just because a specific character is doing better than everyone else by a small margin. In cases like this, the metagame may just need more time to adapt to learning how to fight the character. Time is the only way you can truly find if something is balanced or not.
While the top will undoubtedly receive very small nerfs to alleviate problems in some specific matchups, it is much more likely that lower tier characters will end up receiving several small buffs that allow them to be viable. People are generally a lot more receptive to buffs than to nerfs, and if balance could be achieved solely through buffs, it's all we would do. However, some characters (MK is an excellent example here) are at a level so far above everyone else before nerfs that trying to buff everyone up to MK's standards would be ridiculous, and result in a game that while is not broken in the sense of matchups, is broken in the sense of every character being obscenely powerful. Granted, in general, more powerful characters are more fun to play, as they have more options and therefore more depth, but after a certain point, it crosses the line into "too powerful" and actually results in less depth. It's a fine line, and trying to find it is certainly not an easy task, but I'd like to think that the characters currently considered "good" in brawl+ have not yet crossed it.