I think final timetables are to be released tomorrow on myUnisa, but don't take my word for it.
As far as exams go, all I've heard is that the theory is emphasized more than the use of the API. In broad terms one needs to occasionally remind oneself that eg glRotate is not a "rotation" as your eyes see it, but a matrix that one right multiplies every vertex with, thus changing the four coordinates of that vertex.
Strangely I've found going through the book more fascinating than tough. (Yes it's a bit of work, but not what I'd call a "grind". "Strangely", because this has not translated to programs that work as they are meant to, so I have to then doubt the "understanding" I thought I had.
I think perhaps instead of thinking in terms of "what sections to leave out" it's better to think in terms of "at what depth do I stop". The book tells a story top down. You get the big picture, you dig a bit down from there. You keep doing that (stopping half way in our case). So I would think you need to set out to know the first part like a hypochondriac knows her pulse (as the poet, Douglas Livingstone said, describing the Bateleur). If you get how it all works together, then you can probe down and see if you can better understand one of the more accessible looking details. And so on. Not "what sections do I leave out" but "at what depth level do I call it quits". (That just means "at what level of detail..".