OK, i got my caps and have placed them in 3 different boards. They are;
ASUS P4C800 Deluxe
ASUS P4C800E Deluxe
DFI Lanparty UT NF4 SLI-DR Expert
On the expert i replaced all the 16V 1500uF and all the 6.3V ones around the CPU, RAM and MOSFETs.
On paper the voltage fluctuation of Vcore and Vdimm was improved by 0.1. I was running a Opteron 144 which did not overclock anymore than it previously did

. However, RAM did perform better. At 333mhz (DDR666) the RAM would give 3-5 errors in memtest's test number 5. After the caps were replaced the RAM was able to past memtest's test number 5 with no errors.
Just incase you're wondering RAM was G.Skill's F1 Series TCCD DDR600 at CAS2.5.
In the P4C800E Deluxe i had a 3.0E which tops out at FSB275. After the 16V and 6.3V caps have been replaced the pressy now tops out at FSB280. Which means the 3.0E is running 4.2Ghz on aircooling
The P4C800 Deluxe with the same chip previously tops out at FSB 260, now FSB 265 with no problem.
Both P4C800E Deluxe and P4C800 Deluxe had no othermods other than Vdroop.
Conclusion...
All caps replaced were the same value as the original ones, except in the P4C800 series the 16V 1200uF were replaced with 16V 1500uF.
The caps made quite a nice difference in both P4C800 series. Allowing for a few more precious clocks.
In the expert board it didn't really make any difference, other than the RAM gaining a bit more stability while running DDR666.
Hopefully, i will be able to get myself a scope soon, so i can see why exactly these caps didn't make a big difference in the expert board.