My thoughts exactly. In particular, a lot of the reason for the best process (which is not too different from what the author lays out in this article) is not to enable adoption in the case that the change is a good one, but rather minimizing the damage in the case that it turns out not to be the case. Having a dictatorial leader issue directives that all turn out to be exactly correct is annoying, but not horrible, and probably will NOT actually result in rebellion, as they say. On the other hand, using that method for a decision that turns out to be the wrong one, at least for that team and that project, will maximize the damage. Getting the right process for minimizing the damage of a bad decision is much harder, and much more important, than the relatively simple one of enabling a good decision.