Someone should go find out the transistor counts of each of the major cores. I think you'll find P4 and Prescott have a lot more transistors than Athlon and Athlon64 cores given the same size cache. This is more telling than these simplified statements.
Well I did this myself. Amazingly, it appears they have nearly the same transistor counts...
I always assumed more transistors for Intel's cores because of the larger die size (compare Barton to Northwood). I was wrong there. We still don't know anything about activity factors of these transistors, the optimization of each transistor for speed or power, or the inherent power of the transistor based on the process. The first part is based on the microarchitecture, and the second part is based on the circuit implementation, and the last part is based on process technology.
I'm getting away from my original argument, which was that power consumption is incredibly complex. If it weren't, all these cores would have roughly the same power consumption for a given unit of work, and they don't. So I take issue with the original statement by chipguy that provoked this conversation, and not all of Jerry's additions.