1. Crunch FreeHAL and WUProp always and without fail on every machine.
2. Take the 590 and run PG up to 10M (shouldn't take much longer than a week). [This will get you to show up on the 2.5M, 5M, and 10M MM lists, as the minimum for those is one].
3. Like CBS said earlier, don't crunch everything all at once. Pick one or two projects at a time, and go for them.
4. CPU-wise, I've heard rumblings of Enigma and Cosmology being in scary positions of "maybe we'll close"
5. Pick a MM threshold and crunch every CPU project to that level before you concentrate on anything else. I started off crunching to 10k, then moved on to 100k. If I were you, I'd start with crunching to 10k.
6. Once you hit your 10M in PG, crunch Collatz to 1M (will get you to show up on the 1M list; minimum = 2 in that category).
7. Give Maxwell some of your LHC credits...
7. Take a credit-whore perspective for a bit, and run up some of the high-paying CPU projects to an awesome level. Aqua and Primaboinca are easy to run up pretty high at the moment.
8. Start taking up the other CUDA projects to your goal level.
9. etc...