I think the idea of benefitting the player as well as/instead of benefitting the character is an important one (especially if, like me, you enjoy playing characters that make poor choices, get in over their heads, and/or get their comeuppance at the end). But I also feel like there's a bit of tension with the way that PbtA moves are usually written. Moves rarely define the 6- result, so that's both the worst outcome for the character, *and* the result that gives the player the least input into the story. It makes me want to write a proof-of-concept game in which the player decides how their character suffers on a 6-, while on a 10+ the GM picks a "happy move" to bestow on them.
Viewing post in PbtA Theory: Why You Reach for Moves
I'm reminded of Psi-Run which gives "first say" to the GM when the player succeeds and to the player when they fail--this changes for different aspects of the result other than success and failure, too, which seems like something PBTA-style moves could really easily leverage to shift the texture of different moves. In Psi-Run, for example, while the Goal result is as above, the Harm result gives first say to players when unharmed or minimally so, to the GM when significantly harmed, and back to the player when mortally wounded. Some of the results give first say to the other players rather than the rolling player or the GM.