Any metric that has Wroten, Staukas, and Canaan ahead of the one guy who appears to have a clue doesn't do it for me.
Okafor being awful isn't a hill I'm about to die on, but this is backwards to me. The whole point of analytics is (hopefully) to teach us stuff, not to confirm what we already think. You're free to disregard RPM with respect to Okafor - it's not a perfect stat by any means, but I don't think this is the most solid line of reasoning why.
With respect to Okafor, I think what's going on is largely the Sixers attempt to use him and Noel together. That combo has a mind boggling -25.6 net rating when they're on the court together (about six points worse than Okafor's overall rating). It's just a terrible idea, so Okafor is barely being given a shot to succeed in those minutes. Even if were good, he'd be screwed. That said, he's only spent ~40% of his minutes on the floor with Noel, and he has a raw +/- of -14.4 so far this year. The Sixers have been
dramatically better when Okafor is off the court, and that can't all be blamed on the spacing issues Noel presents.
I wonder what Kobe's secret to a positive ORPM is. He has a TS% of 44.9%, with more FGA per game than Okafor.
This is why BPM is useful. Lets you unpack what RPM is "seeing." Apart from FG%, basically every aspect of his game favors Kobe. He's got a decent steals rate (which helps offense as well as defense), he shoots a ton of threes (spacing impact), he has triple the assists and fewer turnovers. Finally, BPM includes a interaction factor of AST%*TRB%, which basically serves to reward out of position stats (boards from a guard, or assists from a big). Okafor gets credit for neither. He's a poor passing, poor rebounding big.
I'm a little surprised the FG% difference doesn't overwhelm it, but it's hard to eyeball BPM because it's nonlinear.