I'm pretty late to this thread, because I never post anything, but I figure I'd throw in my two cents.
Firstly, I'd like to look at a few very well designed characters, at least well designed from my perspective. The first would be Falco. With Falco, you've got an incredible ground game, and a very solid projectile used for zoning. You can also use this projectile to approach, and after approaching, you can really put on heavy shield pressure if you're technical enough. He's basically got all the things that Melee Falco has. This includes his ability to be combo'd to hell, and his overall glass canon design. Where his match up spread may have been absurd in Melee, the mechanics from Brawl that still exist in Project M, really tone down his overall placement on the theoretical tier list. Is this because of the design of other characters? In part, yes; but the primary reason for his lesser status in PM is due to the actual mechanics of the game. Essentially, what PM is, is a jumble of Smash 64, in that you've got a lot of auto combos/easy follow ups regardless of DI, Melee, which is obviously seen through ground speed and fall speeds, as well as the general feel a faster neutral game in most match ups, and Brawl, where most of the off-stage design of play finds its roots.
It's very easy to point fingers at poor character design, and figure that it's primarily character design that needs to be fixed in PM. In my own opinion, it's very difficult to fine-tune a character, when the actual mechanics of your game aren't completely sound at this point. It's easy to just tweak the design of a character from Melee, and assume that whatever buffs/nerfs you've given that character will make them more balanced. This just simply isn't true. Take, for example, Sheik.
Sheik is no longer a threatening character, at least, not anywhere near as threatening as she was in Melee. With a heavy nerf on her needles and grab game, one would assume that she would still be able to thrive with strong tilts that lead into fair or other great edge-guarding combos. Sure, this is somewhat true against a bunch of characters, but then you look at her off stage game. Sheik suffers off stage in Melee, which is something that could be said for a lot of the cast of that game. She still suffers off stage in PM, with a very short UpB that has a ton of end-lag if you land on stage with it. Basically, if you hit Sheik off the stage, you've got her dead to rights, especially as a character like Meta Knight, or any other character that doesn't have to be afraid to go out and hound someone off the ledge. Now, you've got a Sheik who has a significantly worse neutral game and the same off-stage game as she did in Melee, where PM pretty much demands a very strong offstage game to even be considered a viable character for the most part.
This doesn't mean I feel that Sheik should be buffed in any particular way. Considering the massive changes to the off-stage mechanics of the game in 3.5, I feel that the Dev Team have taken the biggest step in the right direction since this announcement. Once we have sound mechanics established in the PM universe, it will be much easier to determine whether or not a character has poor design.
That being said, some characters stand out above the rest. Doing a relatively simple pro-con analysis of characters can reveal whether or not their flaws balance out their strengths. Let's take a look at Mewtwo.
Mewtwo has an incredible neutral game. He's got tilts with a decent sized disjoint, and very low hitlag. Up-tilt strings into itself, and then into Up-air for a lot of percentage. He can follow opponents DI very easily with float, and he also has a long reaching back-air to cover DI behind him. This is on top of Shadow ball, which comes out relatively fast, especially when you consider the type of projectile it is. Samus and Lucario have similar styled projectiles, but Samus' takes longer to start up, and Lucario's is slower after coming out. Mewtwo gets the best of both worlds for this style of projectile, as well as a 'bouncing' trajectory that can make power shielding more difficult. Mewtwo is a character that, as he is designed right now, can never truly lose the neutral. This is because of his ability to attack out of UpB. Even after getting hit out of the neutral, Mewtwo can teleport in plenty of directions to either go to the ledge (where he can infinitely stall or teleport to center stage and throw a hitbox out) or just return to the neutral and throw out a hitbox to beat out his opponent.
Mewtwo also has a very strong grab game, with a huge grab range, that also extends behind him slightly. He can kill with back throw and up-throw, or down throw into a plenty of moves that will lead to a combo. He can also pressure your shield with well spaced float-nairs. These float nairs can be used to cross up the opponent while they're in shield stun, and then leads to a relatively easy grab while you're behind them, especially against characters with poor OoS options.
What are Mewtwo's weaknesses? The first one is statistical. His floatiness means that he'll die off the top relatively quickly. Any character with a kill throw will be fishing for grabs against Mewtwo at any chance they get. Another weakness is more player-specific. A lot of Mewtwo players are very aggressive with their teleports, and the design of Mewtwo, as was previously mentioned, never really gives them a direct sign that they've lost the neutral. When put into bad positions, a lot of characters can just out space a stupid teleport and punish, which will usually kill. This means that the players themselves must realize when they've lost the neutral.
Considering I'm not a Mewtwo main, I can't really come up with other weaknesses off the top of my head, but I'm sure they exist. What I'm not sure about, and in fact very skeptical of, is whether his weaknesses balance out his strengths. The same concern exists for Diddy, who has an incredible ground game, astounding off-stage game, killing aerials, smashes, and throws, etc. The positives outweigh the negatives A LOT.
Something that does annoy me when reading about whether or not a character is broken, is the negligence of the idea of a skill ceiling and skill floor. A character like Lucas is very, very good, and in a lot of aspects, he needs to be toned down to make sense. A lot of that is being done in 3.5. Most of his design, though, makes plenty of sense. He's a difficult character to use properly, and so you've got to factor in the skill ceiling. You don't see too many godlike Lucas mains, but there are a few that shine so brightly, that when you watch footage of their play, you can't help but feel that everything about Lucas is broken. That just simply isn't true. Looking back at Falco for a second, consider the precision and speed your hands need to properly pressure a shield with no real 'holes' in your pressure. It's very technical and extremely difficult to consistently maintain such solid pressure. A lot of characters require you to work very hard to play well. Characters like Ike, Roy, Pikachu, and Falco are pretty good examples of these high skill ceiling, as well as a relatively high skill floor. You can't just pick up these characters and play them like a god. A character like Mario, however. . .
I'm really starting to ramble on here, but let me give a final summation. Game design is something that is very hard to look at so objectively, especially in a game that's so dynamic as Project M. Right now, the game is still in its youth, and every new iteration clears a huge part of the game's slate. We've yet to see what the final version of Project M will play like, but I assure you it won't be what it is right now, so arguing for slight changes in characters is usually semantics for semantics sake, in my own opinion. Discussing the general mechanics of game play, and sorting those out first, are a priority, and character design, for the most part, should be placed on the back-burner until we feel comfortable with the mechanics of the game.
