Someone always has to come with this particular middle-brow dismissal whenever anything subjective gets posted to HN. Not that I think this is the most brilliant article, but like most things in general-purpose high-level software engineering management, the ideas presented are not quantifiable or specific enough to be tractable for rigorous scientific exploration, you just have to consider them within the framework of your own expertise and experience. Trying to force a data-driven approach on top chaotic human systems where the inputs and outputs themselves are vague and subjective is a quick path to the McNamara Fallacy and other management theory quackery.