Yes, we use diagrams for both CR and RC. The idea is to extract certain key pieces of information and jot it down (very briefly - using serious abbreviations!). The point isn't to take notes the way you would in a lecture but rather to be able to analyze what's going on during your very first read-through.
I obviously can't reproduce both books here, but I'll give you some ideas.
For CR, the most important thing is the conclusion of the argument - the vast majority of questions have one and, for most of the rest, you're supposed to find it in the answer choices. (There are some infrequent, minor types that don't fit either of those categories - but these are pretty rare, so not worth worrying about right now!)
The next most important thing is identifying the type of question (find an assumption? strengthen the conc? weaken the conc?) because that tells you how you're supposed to do the question.
Then, from there, you just want to understand the logic flow of the supporting information - the other information is primarily there to support the conclusion, so how does it work? Where are the gaps / flaws? (Occasionally there's also background information - doesn't directly support the conclusion, just sets the context for the rest of the argument.)
For RC, the most important thing is "the point" or the thesis - kind of like the conclusion of CR. This is the single, major point the author is trying to make through this whole big long passage. Next, understand the purpose of each of the individual paragraphs. You can usually figure this out from the first sentence or two of the paragraph in question. Finally, you do NOT want to get into all of the detail way down in the bodies of the paragraphs.
Sounds funny to say that we're being tested on comprehension but we don't want to learn all those details! There are two interrelated reasons for this. First, as we all know, we don't have much time. Second, about twice as many questions are written for a passage as you will actually see. So there might be 7 or 8 questions written for a passage but you'll only see 3 or 4 of them. In other words, some random details will be in any given question, but you may not get that question... so why bother learning all the nitty-gritty details up front? You won't get half of the questions!
Know what kind of info that passage contains and which paragraph that info lives in - you should have P1, P2, P3, etc. marked down on your scrap paper with notes next to them. Then, when you get a question about some detail, you check and see "oh, right, that stuff was in P2 (paragraph 2); I'll go learn that detail now that I was given a question about it."
That's the high-level idea behind diagramming - get what you need out of that first read-through (but don't get more than you need) and get it in an organized way so that you can do what you need to do with the information.
