Interesting BC76, but I am actually MORE inspired by plugging in to GR2 than going through my rig.
And the best part to recording this way is, I split my signal so I have a clean signal and the effected signal, so I can keep the original performance - and yet have one to tinker with. As soon as I lay down the track, I do apply the effected track to save CPU, knowing I can go back if I don't like it with the clean track.
DigiDis, I do some quick, arpegiated runs and I never experienced a noticable latency when laying down a track - that may have something to do with your setup. If some piece of hardware is sharing an IRQ, or you didn't tweak your machine, then I could see having a problem - but otherwise there is no way you should. I have a older machine - an Athlon XP+ 2600 with 1GB of ram and I can do a lot and stay in sync and not use a lot of cpu.
pgw, I have to scientifically oppose your theory

I usually record my parts late at night, with ambient lighting and a few beers - if you can hear the stress in that, you need to donate yourself to science! Like I said earlier, even when I recorded in a real studio, I still was in the control room when I did my parts, for me, it isn't any different.
I still stand by my earlier comments that only the most trained audiophiles MAY be able to detect a higher-end amp sim, but 99% of people could not correctly identify when an amp sim is being used on a song that is fully mixed. That is my theory!