Water out of the air?

A new invention, called the WaterMill, is promising to allow households to supply their own water by condensing it our of the air. The company producing it, Element Four, is, obviously, very positive on its potential.

For more details on how it works, see The Observer article.

The US currently goes through 30bn litres of water a year, at a cost of approximately US$11bn - while using 1.5m barrels of oil and twice as much water as is actually delivered in the bottles, in the process. If those US consumers could be convinced that the output from the WaterMill was worth swapping for their addiction to bottled water, the ecological benefits are obvious.

What did worry me, though, was this: if the WaterMill was widely used and most houses manage to extract the water they need directly from the air - could this, in effect, reduce the moisture in the air, even further, and therefore reduce rainfall in the areas where it's used? Could the WaterMill actully reduce the likelihood of rain, locally, in the regions in which it is used... thereby reducing the total rainfall in the catchment areas of those regions?

Could this be a zero-sum game, or at least a solution that suffers under a law of diminishing returns?

Perhaps somone more knowlegable on these matters of meteorology than myself could speak to the possibility of such a negative feedback loop being an issue.