It's quite complicated LB testing showed reduced drag for their x-flow profile. And if you can run 10-20mm deeper wavy rims vs normal I'm sure they will be more aero. As we've seen in many materials smooth surface is not always the most aero one.
I've never seen testing done on those wheels, only CFD, and it was rim only. University students also did it as a sponsored class project, which is not good or bad, just context for why it was limited in scope.
My read on this is that stability vs low cda is something of a zero-sum game. Just as an airplane wing generates lift (a lifting force on a horizontally oriented aerofoil) your rim is a vertically oriented aerofoil that efficiently generates side-force when it hits wind at an angle; the better your wheels are at keeping airflow attached, the more side-force they generate before the aerofoil stalls, and when they do, that side-force you were counteracting to keep your handlebars straight suddenly disappears. Conversely wavy rims stall the aerofoil at wind angles where you aren't counteracting so much side force, so the wheel feels stable, but theoretically this costs drag.
I believe wavy rims are a bad solution to the stability problem. The design introduces a lot of parameters to be optimized (and to screw up)--not only aerodynamically but also from structural and production engineering standpoints. I also agree they look pretty kooky and gimmicky.
I think a better ideal is you find a rim profile that stalls more slowly and predictably (against the natural tendencies of air), and the marketing on pretty much every non-wavy rim these days says something suggesting that their wide/bulbous/toroidal/blunt/whatever shaped rim profile enhances stability in this manner. Under the assumption that I can't see aero and correctly identify what such a rim looks like, I think everyone is just better served running slightly shallower and wider rims if you want more stability.
Agree that the time for wavy rims as a "good" solution has passed.
I also agree with the framing of the trade-offs if we consider the shape of a rim in isolation. Rims exist as part of a wheel with tires on a bike, though. In that light, there is not a clear trade between stability and CDA.
For example, you can gain stability and lower CDA by tuning your rear wheel profile to be optimized for the lower yaws it will see. This is the same principle behind the observation that many TT and tri riders have that a rear disc wheel makes the system feel more stable when flow detaches from the front but stays attached in the rear. Following this principle has made the Enve SES wheels among the best for over a decade now (and the 4.5 design is nearly 10 years old now too!). This is the same concept Roval has also been using in their last few generations of wheels. Despite their latest rim profile being shallower in the rear, if you look carefully, they are still following this principle.
Another product that breaks this is the Aero 111 tire. Nearly every test finds that it makes less wind stable wheels more stable while also testing consistently faster in the tunnel, granted, faster in these cases is within the error of the individual tests. I think the reason this works has a lot to do with the top of the tire moving faster relative to the front as the wheel both rotates and moves translationally through the air. With some bikes, this effectively helps with poor aero areas around the fork crown, so it's doing more than you would see in a wheel-only analysis.
My armchair guidance if someone were to want to buy a steel spoked wheelset would be to get the DFS wheels on straight rims with the R2 hubs (since you cant get wavy rims for those hubs anyway). I would pick either their wide shoulder 50mm deep profile for the front or their 50mm deep CX profile (and match with the Aero 111 like I did), and then go for a UV shaped rear. You can see the writeup on my choices here:
https://chinertown.com/index.php/topic,6243.15.html I don't think it's outdated yet, and I believe the rim profiles I used are pretty similar to what CRW ended up using on their second-gen 5060
