Maybe that's true. I don't know; you'd have to be pretty bad at fighting games to look at Melee and not see pretty quickly that Marth is high tier at least. Of course, I hear the early Melee community was, in fact, really quite terrible at fighting games and really struggling to figure out what to do.
The thing is that Marth in Melee was always able to do the things Ken did. It's just the case that no one happened to do them before Ken (that we know of at least; someone could have been doing them in a smaller group that never made contact with the mainstream smash scene). Marth's "metagame" improves, but Marth himself is the same as always. It's not like Marth suddenly got a 3 frame faster recovery on fair after Ken started playing him or anything; that fair was the same as always... as with all of Marth's moves. People can get better with Marth as a group, but Marth himself doesn't change. If we're interested in Marth himself, that's important to note.
A true tier list would be one in which we had knowledge of a perfectly developed metagame for the entire cast and were able to rank them in terms of worth on that scale. Obviously such metagames would still factor in the flaw of human limitation (hence ZSS and PT are not going to be god tier, Marth, Bowser, Wolf, and Mario high tier, and everyone else useless tier), but the limits of knowledge and the biases of popularity should be ignored. Our question is what is the potential, from human players, of the assorted cast. It's a pretty hard question, but we defeat ourselves from the beginning if we're thinking in terms of what people are doing instead of what the limits of the characters are. The experiences of the community do offer us many insights, but we have to recognize the limits of that information, instead of putting the cart before the horse and suggesting that where data lacks we will simply assume weakness until someone "improves" the character.
Matador is pretty much hitting the big point there. The lower on a tier list you go, the more inaccurate you get. I can look at melee's tier list, a list from a far more mature game than brawl, and just notice how my experiences tend to diverge. For the top half of the list, I'm in mostly agreement. I have minor points I don't agree on (like, IMO, Jigglypuff > Captain Falcon and Dr. Mario > Samus), but that's just 1-2 places and such. Among the top 3 tiers, I would not want any moves of real substance. However, then I look at the bottom two tiers, and I have more serious issues. For instance, I think Ness is needing to move up about a full tier's worth. Obviously my melee experiences are already far more limited than m brawl ones anyway, but how would you even argue for or against such a suggestion? A full tier up from where Ness is on that list is still not viable; it's not like there's very much meaningful tournament data for any characters down that low that goes much beyond "horrible matchups with Sheik". With Brawl, we have 7 years less time with the game, only a year and a half of total time really. There's still very wide disagreement over the worth of a lot of characters, and tournament data will simply never give us enough to work with. We have to rely on intuition and analysis to reach conclusions on a lot of this; yes that's basically "theorycraft". The data is a very nice sanity check (if we start concluding Samus is better than Snake, tournament data does a good job of telling us we're doing something wrong). However, it's not an end-all.
I think our biggest questions are ones of viability moreso than exact balance though. I'll just post this list as food for thought since Pierce isn't coming back to this topic. The ordering within the "tiers" is roster order (Mario, Luigi... Snake, Sonic), but they really aren't tiers at all.
Viable: Able to be used alone (no secondaries) in a mature metagame without being a legitimate handicap to the user.
Diddy Kong
Wario
Toon Link
Zero Suit Samus
Pit
R.O.B.
Kirby
Meta Knight
King Dedede
Olimar
Falco
Pikachu
Lucario
Marth
Mr. Game & Watch
Snake
Counterpick Viable: Overall in a sense as "good" as the viable characters, but specific situations (stages, matchups) are just too bad for these characters to the point that they need secondaries. With secondaries, these characters are every bit as good as the characters in the first group and should not be considered "below". What needs avoided is noted next to the characters.
Donkey Kong [King Dedede]
Ice Climbers [stages]
Fox [Pikachu, Zelda & Sheik]
Wolf [several matchups, King Dedede among them]
Ness [Marth]
Semi-Viable: These characters are not so bad that they cannot win in a mature metagame, but they are bad enough that they are an encumbrance to the user. They may or may not depend on strategies that involve secondaries. The degree of encumbrance may vary.
Mario
Luigi
Peach
Bowser
Yoshi
Link
Zelda & Sheik
Samus
Pokemon Trainer
Jigglypuff
Ike
Lucas
Sonic
Not Viable: These characters are sufficiently bad that they cannot be expected to be successful in a mature metagame.
Ganondorf
Captain Falcon
Obviously within each group there are differences in quality, and remember the groups aren't strict quality orderings anyway (Ice Climbers are better than Pit via most analysis methods). However, this is how I have grown to view the game; there are these sort of thresholds of quality after which things really change in terms of viability, and I think we better serve the actual useful purpose of a tier list (informing people who don't know the game what's going on with the balance) better by focusing on that.