I paid extra for a dual hose unit and find myself almost never using the second hose at all. The effect is negligible and not worth the small effort of picking up the second duct from the closet.
And on my unit (trotec) you _can_ close the intake vents with a simple switch if you install the 2nd hose, so its not like all manufacturers are idiots.
How did you determine that the effect is negligible? A single-hose A/C should actually blow colder air locally, because it's using cool interior air to cool its condenser. The problem is, it's also throwing cool interior air out the window, which is thermodynamically insane. Every unit of air it expels must be replaced by hot air somewhere else in the building.
> e.g. not all the compressor's heat being vented outside
This is an engineering problem; the compressor should be well-insulated from the room (thermally and acoustically), and cooled using outdoor air.
> How did you determine that the effect is negligible?
The power bill measures overall how much time the compressor spent running for the same preset, and there's little difference between two hose and one hose mode. I'd presume this is the most effective efficiency measurement I have.
> A single-hose A/C should actually blow colder air locally, because it's using cool interior air to cool its condenser.
Usually it feels as if the air _around_ the ducts/behind the compressor is warmest, rather than around the room's ventilation grills which is where I suspect you would notice warmness if there was a significant negative pressure issue (yes, this is modern France building so I'm forced to keep an aperture to the outside for air renewal, even though otherwise the isolation is topnotch).
> the compressor should be well-insulated from the room (thermally and acoustically),
This obviously makes sense, but it is much easier said that done. People still want to carry the thing between rooms (and/or carry it themselves to their homes) which is why it is a mobile air conditioner in the first place.