While I understand the author's point, IMO we're unlikely to do anything that results in slowing down first. Especially in a competitive, corporate context that involves building software.