Wow, way to bump up an old thread...
In that case, I'm going to add my two cents about this:
Water cooling as it is done in a PC is actually quite the same as air cooling. A copper block (predominantly) takes the heat off the chip, transfers it to a conductor (heatpipes or water) which is then run through a set of fins (heatsink or raditator) through which air is forced to cool it.
The advantages of water cooling are the flexible and variable placement of the radiator and the possibility to control the flow speed of the water, which itself is one of the better heat conductors.
I don't disagree with what you're saying at all
. When I phrased "water" and "air" cooling I was referring to the typical computer cooling setups and not the specific water and air mediums themselves.
Actually, in a typical, SPCR-compatible, air cooling setup, in the only place where the physical mechanisms used are different from water cooling, it's air cooling that uses the theoretically superior
method, not the other way round... The only difference of that kind is that air cooling uses water evaporation & condensation (in heatpipes) to transfer heat from the base to the fins, which should allow to carry a lot more heat than an otherwise equivalent liquid water loop. Every other parts rely on the same physical principles in air cooling and water cooling setups. Of course there are tons of other differences which can (and do) make a good water cooling setup better than air cooling, but it's not about the basic physical principles used - it's about the practical constraints.