Your eRAPM and Easy Player Rating are interesting.
http://thebasketballdistribution.blogspot.com/
It is jumping around, but these thoughts came to mind about your recent work:
EPR might benefit from adding an estimated "defensive usage". I might have recommended that before, to you or someone else.
All 7 of the least helpful players on eRAPM in the playoffs were obtained in the last 25 months.
Win% added per game is another useful format to think about player impact.
A side by side summary of the various stat weights or at least by general stat category for ERAPM, EPR, ASPM, EZPM, Neil's SPM, Rosenbaum's SPM, Offensive and Defensive Rating, Wins Produced, eWins and other metrics would be very helpful if anyone were willing to construct it. It would be hard to get them to an exact apples to apples comparison on everything but it depends on how far you want to push for it.
Which teams did eRAPM retrodict the best and worst this season?
Any overall sense if it does better for guys on middling overall value or the extremes?
Have you looked at any meta-blends with other metrics (beyond the blending already done) and whether combined they are better than either separately?
Have you or do you intend to possibly look at single season regressions and see how the best fit weights have shifted? Is the weight for usage shifting?
Compared to other metrics what is the bottom line on how important individual rebounding stats are or how important rebounding is overall?
I see the top guys historical comparison of EPR to WS/48 and the link to the current season spreadsheet. Could we also see ERAPM and Easy Player Rating for this season side by side, for top players or all players (and perhaps with team identifiers to allow sorting)? And maybe with a few other metrics added too? (And maybe in a CSV file?)