I haven't read Ezra Klein's book, but as far as I can tell Ben Thompson's articles are mostly just-so stories[1]. His model seems to be able to predict everything that happens in tech, and when your model predicts everything that usually means it isn't very powerful at all.
Here's a thought experiment: can you use aggregation theory to make sufficiently powerful predictions to beat the stock market in the tech sector? Can Ben? My strong guess is that the answer is to both questions is no.