Well, if the (admittedly extreme) cases of Brawl+ and Project M are anything to go by, it's certainly possible. Unofficial bug patches aren't unheard of either (dsfix and the unofficial patches for TES games, for example). Typically when a game nears its release, the project enters a code freeze. Planned features that aren't fully implemented tend to get shoved off into a dark corner somewhere, possibly to be completed and tested for a future patch or DLC. Modders can complete these. Bugs that aren't severe enough to block the release are left in the "final" product. Modders can fix these.
As for the community being able to provide better balance than what the developers created with the base product, it's not completely absurd; the modding community has at least four key advantages: the vast majority of the hardest work will have already been done for them, they'll have the advantage of hindsight, they won't have strict deadlines, and they'll have many times more end testers to work with.
An argument could be made that people would cheat, but they're already doing so on consoles. The biggest issues I can foresee would be the added burden for TOs, who have enough trouble getting
official patches correctly installed on the tournament's systems, and the community fragmentation between those who want to give the game time to evolve and those who want shiny patches as fast as they can be churned out. If a balance could be established in the latter case, where people agree on fixing things that are obviously bugs (of which there are surprisingly few; a good example would be Dampierre's game freezing CE) while leaving balance-affecting "bugs/features" (e.g. UJG, GI JG, 66G) in as emergent properties of the game that add to the complexity of the system.
Any action that helps reduce the "developer/end user dichotomy" is a clear win for both parties. Games with protracted life cycles and devoted communities are served vastly better by the bazaar model then the cathedral.
The only way the videogame industry can survive in its current form is to become increasingly exploitative of its patrons so as to make ends meet. Just look what the collapse of the mainframe market did to IBM once personal computing became big.
