There is an entirely missing component to all this power saving logic
1) Faster processing power brings the system back down to idle faster (or turn off)
2) Other components are freed up (memory, HDD subsystem, etc)
So, while it is true that the Q6600 consumes more power at Idle and at 100% load than an E8500 what is also true is that that Q6600 has returned to Idle (or be turned off) while the E8500 is still processing at full load
Next, The memory HDD and other components are also still running at full load on the E8500 longer than on the Q6600
Based on the stock power consumption of the Exbits article, the E8500 consumes 2346 (20*213.3) watts in a 20 minute full load test, the Q6600 consumes 2009 watts during that same test but only takes 16 minutes to complete (16*251.2) , it also finished processing and returned to idle almost 4 minutes ahead of the E8500 system. based on the Q6600 being 20% faster in most tests done
So there is a difference in wattage consumed, in most part becuase when the Q6600 was done so was the entire system, and it was turned off (returning to idle nets similar savings) between the processors, BUT the Q6600 system ALSO was able to stop useage or other components during that time period
Depending on the number of sticks of memory you have in your system, number of hard drives the Q6600 system is much moe economical in terms of wattage consumer per processing task.
Lets not forget what the user was able to now do in those 4 saved minutes out of every 20 ?
Think of 2 cars one with a V6, one with a V8, the big mighty V8 uses a lot more gas on the highway, the V-6 is limited to going only 60mph for a trip while the V8 is able to go 80mph.
take an 80 mile trip and both cars leave the same time, V8 gets there in an hour, turns off motor. V-6 has to drive another 20 minutes to go as far. suddenly the gas savings of using the V6 is not so obvious