I recommend using Mathematica. The student edition was around $130 the last time I checked, and it's well worth it (if you're not a student, the home edition runs around twice that). Formulas are easier to write than in LaTeX, and you have the option of saving as LaTeX, as well as HTML, postscript, plaintext, rich text, and a few other formats. You also have the option of easily being able to play around with the math and see how it works, which you usually can't do outside of math programs very easily.
Mathematica is designed for notes, and there are already many (Mathematica) notebooks out there that allow you to play around with the math, very easily. I also recommend using a paint program running alongside Mathematica, such as Windows Paint Shop or something similar. This allows you to quickly draw a complicated diagram with all kinds of options, such as colors and effects, that are usually hard to do on paper. You can quickly add in pictures into your Mathematica notes, and their are additional options allowing further manipulations of the pictures in Mathematica.
It's easy to quickly make copies of your notebooks and play around with specific things in each copy. It's generally how I take notes, ESPECIALLY IF I'M IN A HURRY.
If you're planning on sharing you notes, you may prefer to use LaTeX if you feel that you're comfortable with enough time to use it. Generally I convert my Mathematica notebooks in LaTeX and then PDF when I'm sharing something with someone that doesn't have Mathematica. This, however, is generally reserved for when I'm going to make an important presentation, and I have enough time to really make the notes look pretty. However, simply converting Mathematica notebooks into another format is usually good enough, or even preferred if the presentation doesn't have to be spectacular.
Memorizing proofs doesn’t really do much for you, at least in the long run; instead, you should try to see what makes them tick. First, what is the structure of the argument? What are the main steps, and what are merely details of carrying out those steps? Many proofs at this stage of your studies have just a single main idea, and everything else is details. Secondly, what kinds of details appear over and over? What basic technical tricks keep reappearing? Those are tools that you want to master for your own use.
Best Answer
It's really an art, not a science. I can tell you what has worked for me:
If you're on page 3, say, and you want to refer to something on page 1, mark it with an equation number or a star or something and save yourself time by abbreviating it: for instance, writing "then using (1.5), we get that (1.1) becomes (1.3)" is a lot quicker than writing out all those equations again.
Directly after each class, do the following:
--Go over the notes quickly and write on the top of the front page, by the date, keywords representing the topics covered that day. For instance, "Poisson's formula", or "Proof that $e^{i\pi}=-1$". This way you'll be able to tell in which set of notes a topic is covered when you're looking for it later.
-- Go over the notes and isolate the things that require follow-up work for you, and put those in your to-do list (you should have one!) For instance, "Understand second fundamental form". Then later, use your resources to take care of these. Do not just stow the notes away and promise yourself that you'll "go over them" later. Unless you isolate specific things that you need to do, they will just pile up and turn into lumps of stuff you haven't taken care of.
Remember that the art of organization is the art of being honest with yourself: what are you really going to go back and do? What parts of the notes are you really going to look over and use later? Are you writing notes with the goal of advancing your understanding, or just because you want to feel like you're doing something?
For general organization tips (and getting the most out of notes is largely about organization), I recommend reading Getting Things Done by Allen.
EDIT: After a number of years, I would now recommend something like Notability for capture, and Anki for retention. Above all, don't rely on your intuition for when good learning is happening. Read the evidence (e.g. the book "Make it Stick" or www.learningscientists.org). Notes are generally low-utility. (Though of course you do need notes to capture information.)