1) "After four years, 36 percent of students did not demonstrate significant improvement, compared to 45 percent after two." Thats 36% of students who spent four years without improving what the study authors deem as important cognitive abilities. Perhaps they learned all of their material sufficiently well to do XXX, but the study says they aren't, in a generalist sense, any smarter.

2) I can't tell from the article whether these statistics were actually influential in determining the students' scores on the test. I totally agree that this doesn't translate well between different majors. If the scores were affected by answering this question or the one about study habits or probably a number of others, then I don't really see the point of the study. They started from some sort of "if people take easier* classes, then they aren't as intelligent" baseline and then showed that people take easier* classes...

* "easier," by their definition, has something to do with writing X pages or reading Y pages.