RealGM

There's a statistic on Hoopdata called APER. The summary of it on their site describes it as following:

APER: Alternate Player Efficiency Rating: PER adjusted for actual Assisted and Unassisted Field Goals, as opposed to using estimates. Also includes Charges.

How credible is it?

Both changes are undeniable improvements to the methods. If that data was available historically, Hollinger would have used them in PER - although since my understanding is that Hollinger used a healthy dose of gut feel in his weighting of components so it's possible once he had the equivalent of APER he might have made further tweaks..

To explain a bit further about the first part: PER makes a rough estimate for the value of scoring vs the value of assists based on how often assisting occurs league-wide. There's zero reason not to get more specific except that Hollinger wanted something that worked historically.