For what it's worth, air with low humidity can be pretty damaging. Lots of materials would become excessively brittle at very low humidities, causing them to break easily. This interesting white paper  suggests that, for most materials, the optimal humidity is between 40-60%.
Still, this is a very cool concept.
When we change one or both of the speeds - the balance shifts, and the effective humidity of air changes to new stable value.
And the more water is returned the faster it ends up in the air. My estimate from wind speeds is that everything will stabilize within 2 days.
i.e. taking water from the air will have no impact. See: http://news.ycombinator.com/item?id=4816580 where I expanded on the topic a bit.
The device in India shown there produces 15 liters of water per square meter per year. The surface area of a bottle is, say, 500 square cm, or 1/20 of a square meter. That would produce about a liter in a year.
http://www.iimahd.ernet.in/publications/data/2005-01-05gshar... talks of 0.042 liter per square meter per day, in extreme.y high humidity.
Clearly, some efficiency gains are needed w.r.t. those designs to effectively make the claim "water bottle tha fills itself". Whether that is possible, I don't know.
just pointing out that efficiency may be better than what you are estimating.
"Zibold's condenser had apparently performed reasonably well, but in fact his exact results are not at all clear, and it is possible that the collector was intercepting fog, which added significantly to the yield. If Zibold's condenser worked at all, this was probably due to fact that a few stones near the surface of the mound were able to lose heat at night while being thermally isolated from the ground; however, it could never have produced the yield that Zibold envisaged"
That made me look for more recent data. That PDF from 2005 was the best I could find.
No idea how long that would take though.
If the extractor area is the size of the bottle's cap (2cm x 2cm = 4e-4m^2) then to get a litre of water (1e3 m^3 air) requires 2.5e6m of air to flow past it - If you want it full in a day that's 30m/s... hmmm... doesn't seem likely.
However if the size of the extractor area is as large as the bottle (say 10cm x 10cm = 1e-2m^2) then a 1 m/s wind may do it.
Based on this I'm not confident in the "marathon runner" claim, but I'm sure techniques for extracting water from wind have been used before, but improvements could be made.
Scaling up the technology obviously increases the viability, yielding something like this: http://www.geek.com/articles/geek-cetera/prototype-wind-turb...
you are calculating the swept volume. but only a small fraction of the swept volume "makes contact" (ie is roughly within the mean free path of a water molecule) with the collecting surface (which appears to be necessary here for water to be extracted).
with netting your argument is closer to being correct. but this is an impervious bottle, not a net.
if you implemented the tube you suggest (and i understand that was simply an illustration to explain swept volume) then pressure would increase in the tube slowing flow and/or flow would be predominantly around the edges, with, again, a stagnant higher pressure region in the centre.
If there are pollutants in air they will be in the water, but it's not a actually a problem because, although they might be bad to breath, most air pollutants are harmless when eaten.
Carbon (soot), Ozone, NOs, sulfur dioxide (smog), etc are harmless to eat.
I'd imagine it'd have to do with the digestive ecosystem, if there were any that worked that way.
But if there were it would be something that requires acid or digestion to "activate". Or perhaps something that can't be absorbed without something the only exists in the gut.
Bacteria actually could do that. Inhaled they would rapidly die, but in the gut they could live (more food for them) and cause illness.
Also, the gut will absorb things faster than the lungs. So I could easily imagine a toxin that has little effect when inhaled (since it's absorbed too slow to cause damage) but causes trouble when eaten (everything is absorbed at once).
But I don't know of any specifically.
Any air that distills out water that was worse to drink than what was available locally would probably kill you to breathe just as quickly.
But imagine that, instead of a special water bottle, you have a poncho-sized sheet of coated plastic in the shape of a windsock with a coiled wire ring at the opening and a pouch at the end. If this could work like the beetle's wings and take water out of the ambient air, that would be interesting.
What did work fairly well was a transpiration bag . It took a few bags to get a meaningful supply of water and one needs to be careful about the foliage used.
You could create condensation with only a light breeze, but it probably wouldn't yield a significant amount of water. I think the best case for these would be to attach them to balloons and put them high into the atmosphere (where it's much colder), so that they could produce even more water with the same amount of energy.
(MIT, 2011) "In some field tests, fog harvesters have captured one liter of water (roughly a quart) per one square meter of mesh, per day. Chhatre and his colleagues are conducting laboratory tests to improve the water collection ability of existing meshes."
I find it curious that people who design nifty new devices are regularly referred to as "scientists". This is not science; it is engineering. Why not call it that?
That sounds pretty hard, but it would be a safe bet to say you could buy one within fifty years. Maybe longer if you want something nicer than pure ethanol.
 http://2011.biomimicrydesignchallenge.com/gallery (#38 -- 5 rows from bottom in center)
I realize that extraction rates are probably slow right now, but imagine this technology n years from now.
This sounds like some sort of nanotech approach that causes the water to condense without needing a temperature change. It may not even require any power, if you can accept a slow rate of water collection.
Note, I said: if we do this on a massive scale.
We have these enormous oceans on earth, and if the air gets dry more water will simply evaporate from the ocean. It takes about a month for air to completely circle the globe - and a whole lot less than that for air from any particular place on earth to reach the ocean.
So even if you completely dry out air in one spot it'll be replenished within a day or two at most. Meaning that overall you never will actually dry out the air.
And remember that on a global basis water is more or less never created or destroyed. So any water you capture from the air, will just end up right back in the air.
You water a plant and it converts the water to hydrocarbons, those hydrocarbons are then eaten, and are "burned" and are converted right back into water.
Same for the water you drink - it's only in your system temporarily, it will end up in the environment again later.