[azernik]

As someone else noted, this system functions by using a refrigerator to keep air in the condenser at the dew point; the power required to do this varies based on humidity (changes the dew point) and temperature (changes the temperature gradient your refrigerator has to maintain between the dew point and the outside temp).

Chilling below the dew point doesn't add much, if any, performance.

(And yeah, that kW = kW-h convention hurts my brain.)

[08-15]

Dehumidifiers are not a new thing, I have one right here I can observe. It either runs or it doesn't, so it either consumes 200W or it doesn't. It doesn't regulate the temperature of the condenser, the condensing water does that quite nicely. Since the machine is switched by a hygrostat, you can surely say something like "0-200W" on average, but then it also has a capacity of "0-10l/d".

Power consumption doesn't depend on humidity. Capacity does. So if they quoted "0.8-1.8kWh/l", that would make sense (but be ridiculously low). If they quoted 1.8kW and 40-100l/d, that would make sense, too. But 30gal/d and 0.8-1.8kW implies that the machine throttles the compressor in order not to exceed the advertised capacity, which it doesn't reach under real world conditions anyway.