I don't know about that 2.5-4-4 mushkin RAM tightening up to 2-2-2-5 timings and ontop of that OC up atleast 20MHz? Seems pretty unlikely only because getting the ras to cas and the ras precharge down to 2 from 4 seems like the RAM wouldn't at all handle it, muchless after OC'ing the actual frequency....Leonidas already fried his RAM trying to do the very same thing. OC the hell out of the actual RAM frequency and then tighten up those timings ontop of that, although he tried changing the Cas latency from 2.5 to 2 is what really caused his fry I believe.
2.8v 2.5-2-2-5 at 235MHz is one insane OC for something that's stock 2.7v 2.5-4-4-(Not sure) 200MHz don't you think? I have very little RAM OC'ing experience though, but I know the RAM I'm workin with now definitely isn't that good.
Infact, I forgot, I removed my 256mb of PC2700 because I gave a mobo, that stick of RAM, and a 9600XT 256mb card to my friend..not GAVE, rather sold it to him very cheaply, but yeah.
So now I got my FSB at 202 just to have it at 202 instead of 200
but it's running 2.5-3-3-7 opposed to it's stock of 3-3-3-8. I've tried changing it to 2.5-3-2-7 but it didn't like that 2...and then the last time I upped my NB voltage just by .1 after I rebooted my RAM was fried for some reason and I had to RMA it =/ It was running at 215MHz at the time on stock timings though so I don't know why raising the NB fried it in that instance =/