Well this was interesting and may prove to be helpful.
There are a couple oddball pieces of advice here.
One is that undergrad research experience is irrelevant for admission to grad school. I have pretty consistently heard exactly the opposite.
The other, which I consider completely bonkers, is the advice not to study for the Math GRE. "A person intending to go to graduate school in mathematics certainly should score *at least* in the high 700's on the advanced mathematics subject area exam. ... If you have a solid background, and if you get a good night's rest before the exam, then you should do fine. ... I can see no reason to study for the Math GRE, and I do not encourage you to do so."
Right... Well, since less than 20% of students who take the test make his high 700's minimum, perhaps the rest forgot about the good night's rest beforehand? It seems that his assumption is that students applying to grad school all have good backgrounds. And, yes, for students with strong undergrad backgrounds, the Math GRE is pretty basic. But there are plenty of students with weaker backgrounds who NEED to fill in some gaps in their knowledge. To be honest, probably the large majority of math majors, even ones with their eyes on grad school, have "weaker backgrounds." And, regardless of background, anyone who doesn't at least work the sample exams for such an important test is just crazy.
I think this attitude may ultimately come from propaganda by testing companies ("Our test is magical and really tests innate ability, so it's impossible to study for it!") and seems to have been prominent a few decades ago.