Sure, the article is about about pursuing science rather than thinking about it. But that's the author's whole point. Enjoying a career is all about enjoying the day-to-day work: If you love thinking about DNA but don't love pipets, you're going to be unhappy a lot of the time, because life in the lab is about 10% deep thought and 90% pipets. (Or, in the semiconductor laser lab: 10% deep thought, 50% misaligned optics, and 40% mysterious process problems that you will never entirely understand, but which you will eventually solve by spending months on end turning knobs in a strategic manner.)
Same general advice applies in engineering: do your best to attach yourself to a lab, and see if it catches your fancy. Best way to test-drive a career choice.
He's kind of down on textbook-and-problem-set coursework and large lecture classes. This is not universal. Some large lectures are large for a reason -- the professor is a star. And some textbooks are really good, and some problem sets are worth sweating over.
Which is not entirely a useless endeavor; if you don't know about what has already been discovered, how can you build on it and go further? How will you know what has already been tried?
That being said, I agree with the sentiment. Most of what we teach undergraduates is about the knowledge science is produced, rather than about the process of doing science itself.