I think the making sure the water doesn't 'stick' to the outside of the walls is key. I think that's exactly what they're trying to do. I read something once about charging both sides of a metal plate when actively cooling it. The idea was to get the 'skin' of air to move accross the plate rather than sticking to it, at a molecular level. Apparently this worked really well - but it was a sort of non-comercial research thing.

I'm confused about the tubes - in the HWLabs link I posted they clearly said that the mixed metals problem was resolved by the nano-coating. But Skinnee says they're not alu in the water path so I guess I sit corrected