Back to the MOSFET issue ... the major reason the Prescott thrashed the MOSFETs so hard is that it used a lower voltage and consumed more power. Since P = V*I, the current requirements went up significantly. The power dissipated by a circuit is P = I^2 * R, so as the current increased (and the resistance of the power supply circuitry remained roughly constant), the power dissipated by the circuitry went through the roof, resulting in high temperatures. The resistance of a MOSFET (and the connecting wires) increases with temperature, so the power increase will actually be slightly more than what the simple formula above gives. Incidentally, this is why MOSFETs can be run in parallel and bipolar transistors can't. The Ron ("on" resistance) of a bipolar transistors decreases with increasing temperature, so if one heats up slightly more than the others, it will carry more current, so heat up more, so carry more current, etc etc. Typical thermal runaway effect. However, since Ron of MOSFETs increases with increasing temperature, if one heats up more, the resistance increases so the others will take more load, and things all balance out. OK, back to the topic ...
With the 3.41 revision of the "AMD Athlon 64 Processor Power and Thermal Data Sheet" AMD gives the current requirements of the Winchesters: IDD max = 45.8A. For a Newcastle or Clawhammer, you're looking at 57.8A. So in fact the power lost in the regulation circuits will DECREASE by about 37%, so the MOSFETs will run cooler.
Incidentally, the power requirements for a FX55 were released as well, and that chip is a scorcher.
A FX53 gobbles up 57.4A at 1.50V, with a TDP of 89W. A FX55 eats ***67.4A*** at 1.50V, with a TDP of 104W!!! That's the hottest AMD CPU to date, AFAIK. And compared to running a Winchester, the MOSFETS are going to be putting out about 116% more heat. So it's the FX55s you've got to watch out for, not the Winchesters.