> Well, how about producing highly skilled teachers, that seems important in a technological society, where mathematical skills are in general critically import

> Well, how about producing highly skilled teachers, that seems important in a technological society, where mathematical skills are in general critically important. Hence, a critical part of the PhD student's job (and the professor's) is the instruction of the undergraduate population. This article doesn't even mention the importance of that role. They don't mention it because it's unimportant. Your thesis committee cares about your research and that's it. If your advisor is well funded, you can easily finish a stem PhD and never teach a single class. If you do have to teach, getting perfect feedback or terrible feedback from student evaluations has zero bearing on whether you earn a PhD.