Yeah, that's a good question. Federal VOC regulations require that all modern gas tanks be under moderate pressure. Excess pressure is vented through a canister filled with activated charcoal, which does a great job of absorbing the volatile organic compounds. When "make up" air passes back through the canister, the VOCs return to the tank. In this way the canister has a long life of absorbing and desorbing VOCs. Water vapor isn't much slowed down by the charcoal, so it can go into the fuel tank also.
During re-fueling, warm moist air has a great opportunity to enter the tank and condense into the cooler fuel. At this point, only a water separator (density separator) will remove the water, unless you add some ethanol to absorb it, but, of course, ethanol has its own problems...
Car manufacturers learned a while ago that it is better this way than trying to put a self-regenerating water absorber in the fuel line.
The one in your compressor is either amorphous silica or a refrigerated heat exchanger/condenser. Once it is saturated with moisture it stops working. Silica gel has to be regenerated by exposing it to high temperatures to dry it out again (silica gels usually have an indicator dye which tells the consumer when it is time to regenerate). The condenser has to be purged of water every once in a while, usually with a trap solenoid on a timer.
Neither of these will work well on a automobile or marine application, because of the regeneration step. With silica gel, regeneration would mean high heat through fuel lines (!) and a condenser would be an energy hog.
So, until someone invents a perfectly efficient, self-regenerating, cheap desiccant system, I think we are stuck with water separators and absorbers.