All I see here is a bit of advice on what to measure, and nothing on how to measure it. The problem with measuring software engineer performance has always been in the how, not the what, so this article is just noise, IMO.