My wife saw the video of this on Facebook today. I have no idea whether this concept would be economically feasible for large-scale development. After a Google search I found the following Time article which provides more of a description. Thought I would share this to get some feedback from the broader community.
I'm sure I recall reading something about this before ages ago - can't say I read a lot of details - but I don't see how it would be particularly energy efficient. It would work like an air conditioner or dehumidifier trying to condense atmospheric moisture onto the cooled condensing surface and taking away the water. That means having to use energy to pump the working fluid around the loop (similar principle to a fridge).
I'm curious though - is the energy input require to condense the water comparable to that required to boil it? Does it count as a phase change to take water vapour that evaporated and to then condense it again? If so, I think the energy requirements would be comparable to boiling sea water to condense the steam to get fresh water? And that assuming the rest of the setup was comparably efficient (questionable).
Even if energy wasn't a problem - over a larger scale - I could see problems scaling it up - after all - how much of the local water vapour can you remove before you need a fresh supply of air with new water vapour? How do the maths look for amount of water you can extract?
On the whole I would have thought passive collectors (if such could be designed - to condense the fluid at night and let it run down to a collector at the base) would make a lot more sense - but cannot see how anything like this would provide water in the volumes required for industrial or agricultural use. Short term baseline survival - perhaps.