>>13958140I'm not that anon, but it depends a lot on the faculty and the course. I'm doing my masters in engineering mathematics, but I've taken several courses in the more "pure math" side of things as well.
One big factor is that in technical universities mathematics serve as a way to "filter out" underperforming individuals. So what they test you on the exam (like
>>13958014 explained) isn't necessarily mathematical understanding, but rather cognitive ability, since understanding math they're not going to use in the future is less important than being able to learn whatever their future employer wants them to learn. This is very much the case for the entry level courses, calculus, linear algebra, etc that are mandatory for most STEM programs.
On the flip side, you'll find that courses intended for more "research-oriented" programs are the exact opposite of this, where they don't care at all about how many assignments you can solve, rather how well you can solve them. These are usually the proof-based courses like real analysis and functional analysis. And even then there's a lot of variation between different faculties, some faculties want to produce more math PhD candidates, they focus more on understanding, while others focus more on wringing out as much juice from each student as possible.
Because, to answer your question, drilling procedures doesn't necessarily make you understand it better or faster. You'll notice several math PhD candidates who had Bs and Cs mixed in with their grades, because while they didn't perform as well plowing through exercises and solving problems they were better at the more abstract and deeper understanding of the topic
