Honestly, I'm not all that interested in it, and am certainly unqualified to suggest improvements to the model. I'd have to go back and read the article more carefully, along with probably a bunch of other material as well, to evaluate it properly. Seems like one of those scenarios where if you torture the data enough it will tell you whatever you want to hear.

It just struck me that both the red and blue traces in the last graph he presents, where he arbitrarily(?) adds another term to the math, are so ridiculous as to not be worth further consideration. If there's a more subtle point, like I said, I'd need to re-read to find it. Either that, or the author needs to bring his lede up front and center. Demonstrating that one bit of math is bad by making up another equally bad one doesn't carry the point as effectively as he was probably hoping it would.