The necessary data is only available from 1996 onwards, hence no old heads.

I didn't go into a deep dive on the methodology but the broadstrokes seem reasonable: how much impact did each guy have on play-to-play win probability?.

The big problem that comes to mind is that an off-ball master like Steph or shot deterrer like Wemby will be undervalued because they have large impacts even when they aren't ultimately in the play. Maybe that's alright, since we don't know if they would have actually made the play that the other team was afraid they'd make.

Depending on how win probability is estimated, there could be a pricing in effect that would also understate how clutch great players are: i.e. Lebron's win probability should always be higher than everyone else's because we know he's more likely to make great high-impact plays.