Okay, a couple of thoughts.
When people say "water-cooling", they generally mean liquid cooling, active or passive. Water would, of course, be a poor choice, in as it's conductive (well, not distilled, but this isn't a lab setting) and it boils too readily. Sure, it could be played with, but why bother. Vegetable oil, for instance, is non-conductive and boils higher. People make very crude passive cooling systems by submerging a (usually heavily-overclocked) motherboard directly into a pool of oil, the cables just snake in. The oil in the center gets hot, moves to the top, is cooled, and moves down the sides. Pure convection. Anyway point is it's nonconductive, and something along these lines could easily be used. I'm sure whatever Cooligy uses is similar. But in a closed pipe this small, there's really no reason to worry about leakage, in a non-phase-change system, because we don't really need to use pressure. Plus, Cooligy uses a different type of pump, that's more in line with ion propulsion than the typical mechanical pump (and compressor) that is likely to blow in your car or home heat pump.
On to cooling pipes. These aren't as effective as an active liquid cooling system, and i think Apple was just experimenting with the G5s. These are more like the PowerBook version of a liquid-cooling system, assuming they don't want to use a Cooligy system in there. It has it's uses, but there are more effective methods that are just as safe and reliable.
paul
