honestly, it's a pretty good approach. I know people like to think that anything that isn't perfect is automatically garbage, but that's not the case. he's taking a lot into account including offensive, transition, and defensive ability while trying to apply context in the form of difficulty of minutes. and he was looking at all kinds of things for each of those 3 branches. The thing is, the list still looks wrong, and he even admits it. that's because there's still quite a bit that can't be quantified so easily, such as the impact of systems, teammates, and individual physical attributes. So to me it's not that he's using too much irrelevant data and it's all garbage, but that he's not quite using enough.
I also think people are putting way too much faith in a player's reputation and automatically disregard any evidence to the contrary, no matter how strong it is. I know people still think that points are the purist form of talent evaluation and looking at anything else dilutes it, but I think the opposite is true. The more data you can look at, the better