I think this is a difficult question to answer usefully because the notions of "fun/enjoy/..." vary so wildly among individuals. There are also extremely disparate impulses about delayed gratification, long-term versus short-term this-and-that...
Nevertheless, especially after the questioner's edit that indicates that the context should be "high-end" programs:
First, I think there is considerable utility in realizing that some difficulties are genuinely mathematical, and some are essentially completely artifactual (even if traditional!) created by curricula or instructors. Unless we go so far as to believe that mathematics itself is altogether nothing more than an artifact, which I think is too extreme, it can be important to learn to distinguish genuine complication from contrived. This is not to claim that contrived issues needn't be fun or entertaining or profitable. Indeed, doing well in contest-math, if one has the inclination and the capacity, is effective PR and can be an ego-boost. Doing well in homework assignments or exams can have similar benefits. However, I think one should not lose sight of the fact that such things are mostly quite deliberately contrived by other people as challenges to others, without regard to deeper meaning.
This distinction might seem pointless... but I have seen many cases of failure to appreciate the distinction between genuine and contrived lead to vast wastes of time and effort in misguided attempts to defend against sniping and prank questions. Sure, prepping for contests may warrant such, but at a certain point one should realize that it is not only essentially impossible to pre-emptively defend against deliberate prank-questions, but it is very expensive... and displaces much more worthwhile work.
I claim that this argues for moderation in line-by-line reading of texts, for moderation in doing exercises at the end of the chapter, and for moderation in thinking mathematics is only legitimate to the extent that it is formalized.
Rather, most mathematics is based in physical intuition (if of a refined sort), and one can come very close to "correct answers", at least if one is allowed to work iteratively (which also seems disallowed in most formalized settings), without necessarily playing the game of formal arguments. I'm not trying to argue against formal arguments per-se, but only that formality as an alleged minimal criterion for mathematical correctness ought not be a foregone conclusion.
I mention this sort of thing as a push-back against the (understandably convenient) formality of most texts and most curricula...
As an action item, I strongly recommend working to get an overview of the mathematics that is reflected in the standard curriculum, since that larger picture often explains the quirks of small details, thus avoiding any seeming need to "memorize" formal things whose motivations would only arise later. That is, read lightly through textbooks... just to see what happens later, how things re-appear or don't, etc. Especially beneficial is the bit of psychological familiarity this confers, in advance of appearance of new ideas in classes. Many people greatly underestimate the role that a sense of prior familiarity can play in "comfort". Discomfort is obviously an inhibitor, for most people...