The Results
Firstly, a huge thanks for Datamonkey, who called over and taught me so much about what actually affects what in terms of voltage changes and turning off redundant stuff on my motherboard. For the knowledge alone it was worth the money.
So ... moving on, what actually happened? Well, the bad news is that the P5N32-E SLI is a crap motherboard for overclocking. We found it impossible to keep it stable at anything above 2.7GHz (300MHz FSB, x9).
The good news, on the other hand, is that my processor is excellent. It's running nice and cool at load (about 42C) with stock voltage of 1.225V. My memory is also great too and the power supply is top notch too. It seems that the only duff component is the motherboard.
It's an awful shame really because the board itself is very easy to set up and use and recovers quite well from crashes. A few times when we were running quite high NB and SB voltages it stopped recognising the IDE bus, and there was also the time when we upped the FSB to 333MHz and it stopped posting (which required a jumper switch change) - but other than that, it was pretty good.
Looking at most other posts on the subject, using the MB without SLI, it's possible to go pretty high - although this is the exception rather than the rule. With SLI, the next highest report was 2.93GHz. Not good.
I'd like to hear from others at this point about the P5N-T and SLI. I'm going to do more reading today on the subject.
