This is with a ton more tweaks. Most improvement was gained by maxing out my CPU/NB multiplier. I also got a few hundred MB/s by adjust various sub-timings, the biggest gain from tWR and tRTW both of which went from 8 to 4 clocks.
Ok so I am very close to the max I can go due to the FSB maximum. I am fairly sure i could raise the CPU/NB more if there were more multipliers available to me.
I did manage to get to 5-6
-5-15-20-1T, I did that with a small voltage increase.
Unless someone else has some other ideas I am going to say this is the end of my ram timing adventure. Time to move onto pushing the HT Link to get more performance out of my gtx 590.
Thanks for everyone' input, xd_1771 especially.