That's an interesting point, and it makes me wonder if there's a level of nuance to be found here. For example, animations are acceptable iff they do not extend the time required to perceive the requested action.
In other words, it's already going to take me some fraction of a second to perceive any change; animations within that fraction of a second are perfectly fine. Anything that extends the change past that fraction of a second, however, is eating into my productivity (or at least, so the author would claim).