Wind tunnel testing is cost effective for fine tuning the CD, especially where CAFE penalties are at stake and MPG on the sticker helps make sales. RVs are exempt, there are not even any test standards.
As happened with cars and trucks, the first big gains can come from reducing frontal area (get rid of basements, high ceilings, roof-top add-ons) and you don't need expensive wind tunnel time to measure results, the answers can come from coast-down tests and towing with a Tapely drag meter. A few manufacturers do measure drag vs speed with full scale RVs, both towables and motorhomes.
Wind tunnel comes in when you have the big stuff like size and shape sorted out, and you need a close look at air flow to take care of things like blending in mirrors, door handles, window to body transitios, hiding wiper blades from air flow, etc.
Basic shapes can be developed using small scale models rather than taking full size vehicles to the wind tunnel, lift and drag on bare models scale pretty well.
We've had a few rounds of trying to market sleeker RVs, and some of the B manufacturers have done pretty well with aerodynamics, making smoother transitions on basic shapes, blending in some of the high drag add-ons we insist on having. I think they are more successful in the market with the space compromises because B buyers have already decided where they want to be on size vs space.
C's could get better on the new front-drive, low-floor platforms becoming available, although we'll probably see Anmerican size motorhomes on the platform, the lower and narrower Euro-size will have to carefully test the market.
We've also always had a fewer lower profile, more smoothly shaped TT lines, but buyer preference remains with the cheaper to build square boxes that maximize usable space inside, and 8-wide, full height TT still sell better than the few 7-wide low profile models available. People want more headroom, more ground clearance, more space.