As a ball park number you could assume that - in arid areas - 1 m³ of air has less than 10 ml³ of water as vapour. That would be a minimum of 100 m³ (a cube with an edge length of roughly 4.5 m or 15') of air for 1 l of water even if you harvest all of it.
You also need to extract and disipate 2.26 MJ/kg to condense the water (more at temperatures below the boiling point).
So I am a bit skeptical too.