>Our estimates of the total surface partitioning capacity are much larger than if the reservoirs are taken to be thin organic films on smooth, impermeable surfaces.

... so is "smooth, impermeable surfaces" the current begrudgingly-accepted model or something? because there's no way any person who has ever been in a house would think that's a reasonable model. permeable surfaces are all over the place, literally most of the place because it includes essentially all walls and therefore wall interiors. managing that for e.g. humidity is a significant part of building design because it's completely inescapable... and that's before even touching stuff like fabric where your average couch probably has more surface area than all structural surfaces combined.