Perhaps because everyone here knows there's very little actual, proper science on programmer productivity, and even less of that is applicable in workplace environment.

There is no break from the pattern in this thread. People still call for following the science - they just consider this article's science to be the RF engineering equivalent of Wakefield study.