I think we're almost to the point where discussing character attributes should be superseded my discussion of MU spreads for the purpose of tiering. Knowing that which characters have desirable attributes (mobility, kill set-ups, OoS/defensive options, etc.) provide reliable heuristics to their eventual placement. But at the end of the day, their (weighted) spreads are the only things that matter for determining their viability.
FWIW I have an excel sheet of the top 10 Brawl characters and (my opinions of) their MUs which I used to calculate the "top tiers" off of weighted spreads.* The most valuable part of this project was not the final result, but the fact that it allowed me to test how the top tiers changed across various plausible scenarios. Sometimes I found that my intuitive tiering of a character lined up with a particularly optimistic/pessimistic MU spread, in which case I knew that I was probably overrating/underrating that character. For instance, I found that I had to be very optimistic about Falco/ZSS and very pessimistic about Snake/Pika if I wanted to place them in their conventional spots. I concluded from this that either the conventional MU or tier list wisdom had to be wrong.
I'm thinking of doing the same thing for this game soon. My methodology requires that the relevant threats be identified beforehand, but we've already done that through tournament results + the aforementioned tiering heuristics.
* = the weighting in this case as fairly simple. Everyone agreed that
/
ran the game, so I just doubled the value of those MUs relative to all others. I'd argue the weighting procedure for this game could be just as simple...anyway, the tier groupings were just based off of obvious gaps in the clustering of the final numerical values.
And FWIW here's what I came up with for Brawl, which you're welcome to use to judge whether or not I'm a total crackpot:
EDIT: WOW I can't spelling and grammar today