I'm going to go out on a limb here, and assert that this discussion is conflating two different issues: mechanical balance, and narrative balance, which mirror the concepts of task and conflict resolution, if you're familiar with them.
Task balance is the standard idea that every in-world character concept should be of equal in-world effectiveness.
Narrative balance is the idea that every player should have equal narrative input over extended periods of play.
----
To assist me, I'm going to introduce two different kinds of gaming mechanics: conflict resolution, and task resolution.
To go back to the *BANG* You're dead! scenario, conflict resolution is a game mechanic that occurs ONLY when players disagree. In this case, they might, say, have a bidding war with narrative tokens to see who gets to decide what really happens.
A task resolution system, on the other hand, is concerned with modeling (at least in the abstract) the imagined world. In this scenario, a task resolution system would model how good Character A was with a gun versus how strong Character B's body armor was, and assign an outcome based on that.
----
This whole argument really boils down to one of the great failings of task resolution systems: if the task resolution is task-unbalanced, then narrative balance suffers (some players dominate the narrative compared to others). However, building a fully task-balanced system narrows the range of viable character concepts to those that can be modelled at approximately an equal level of effectiveness.
Conflict resolution systems dodge the bullet addressing narrative balance directly. However, these kinds of abstract/meta-level mechanics aren't to many people's taste.
Some people have brought up the Buffy RPG, which is a good example of a hybrid system. While fundamentally a task-resolution system, it also has a narrative-balance regulating subsystem.