Seems better than just learning to hate math. Our current system is certainly not effective in teaching young people to add fractions, which I see as the single greatest indicator for competence in college mathematics (sigh).
Ok, after another moment's thought your argument seems similar to "Why should kids read stories and books? To succeed in writing dissertations they need grammar. All that extra reading of literature should be confined to the side..."
It actually sounds a lot better.
> If students are going to succeed in his college-level courses then they need the basics down cold, which means memorization, repetition, and application to concrete problems
Further, the only time I ever needed it "down cold" was when college professors were insisting I not use a calculator on exams. The rest of the time, I was better off knowing about the context and motivation behind what we were doing, not calculating faster.
This experience is shared by friends of mine who are now pursuing PhDs in the hard sciences or working in finance: the computer does the calculating, but you need to know how to write mathematics, not how to manipulate numbers (by hand).
An occasional lecture on a 'fun' topic (or even better, an assigned reading and short essay -- expository writing should be practiced in every subject starting in middle school or earlier) could be motivational. But the author wants to spend "20% of class time opening students' eyes to the power and exquisite harmony of modern math," which he expects would "feed their natural curiosity, motivate them to study more and inspire them to engage math beyond the basic requirements." That's fantasy land. The author sounds like what he is, a UC Berkeley mathematics professor. If students are going to succeed in his college-level courses then they need the basics down cold, which means memorization, repetition, and application to concrete problems. That is, the "stale and boring" stuff.