Postings in response to a conversation regarding the most optimal rear disc selection for a Cervelo P3C--
[Cervelo] did say this sometime last year; the gist of [their] argument is that frames that do *not* have a well-designed rear wheel cutout do better with the toroidal wheels and that the Ceverlos (with their good seat tube design) work very well with flat discs.
I've tried a number of different types of discs on my P3C--best performing, with no doubt in my mind, is a lenticular disc with a narrow tire; second best is a flat disc. The toroidal discs tended to be slower. I'm confident enough in my results that I'm using an older Hed lenticular disc for now while I wait for a replacement for my Comete (which I totaled last season).
The toroidal wheels--seems to me--mostly benefit those who want to run wide tires (remember, there are two edges that need to manage airflow on any wheel) for reasons of rolling resistances and comfort. As far as I see, for lowish yaw angles that faster riders normally see, narrow/flat=faster. The main caveat to this is the fork and seat tube interaction with their respective wheels.
Now, you have a frame like Look's that makes a whole different argument regarding rear wheel aerodynamics; they take into account the possibility of an air dam being created between the tire and seat tube of a rapidly rotating wheel. Their solution is to move the wheel *back* from the seat tube--[Cervelo] indicated that this is a good solution for frames that don't have a good seat tube design.
There have been studies that indicate that while body position is probably the most important component for aerodynamics, when looking at TT as a system, the body itself really doesn't impact bicycle or frame aerodynamics in an appreciable fashion. In other words, you can put two people on the same bike with very different morphology (very stocky, very thin) and the frame/wheels will perform the same. In terms of magnitude order, you are completely right--a good aero position will be faster than a bad aero position, but both positions on the same bikes won't really impact the aerodynamics of the bike itself. (Interesting note--even given the same levels of fitness and watts/kilogram, a thin rider isn't always going to be the quickest, based on aerodynamic factors--some "rounder" people actually will be able to manage airflow better than their thinner counterparts--think about this one for a bit!).
So, when one considers aerodynamics, one needs to consider both the overall impact of the system as a whole, and the individual impact that individual equipment choices might have on the performance of the system. Rolling resistance is important--no argument here. But rolling resistance is a bit like the impact that an aero position has on watts production. For many riders, an effective (nb: not necessarily the most aero) aerodynamic position might yield less watts at FTP, though the riders' times will be faster than their standard road positions at higher levels of watts. Rolling resistance, in other words, needs to be balances against a tire/wheel combo's aerodynamics--sometimes, its better to give up some rolling resistance gains for even greater gains in aerodynamics.
Here are some examples--the Hed 3 is a great all-around performing wheel. The H3 is optimized for 18-19mm tires; throw a 22-23mm tire on this wheel, and you'll have greater losses associated with reduced aerodynamics, than you will have gains from the better rolling resistance associated with the wider tire. Easton's 90mm TT wheel is a perfect match for a 21mm tire; testing with a 19mm tire increase rolling resistance and decreased aerodynamic performance.
My Mavic Comete last season and this season's Hed lenticular both have narrow rims; a 19mm tire worked best with both, though there wasn't a huge difference when using a 21mm Vittoria on either of them. The key, it appears, is that airflow has to be managed on the front of the wheel (it's leading edge) and the rear of the wheel (its trailing edge). On my P3C, I can get a wide tire to draft the seat tube, which addresses the leading edge problems associated with a wider tire/narrower rim, but the trailing edge was problematic. If I was using a toroidal wheel (Sub 9, etc.), I could hide the wide tire behind the seat tube, and then let the toroid bulge of the wheel's rim address the trailing edge aerodynamics. In this situation, the wider tire would give me better rolling resistance, while the bulge rim helps minimize aerodynamic penalty. But given the extremely close clearances of the toroidal rear wheels have in my P3, I find best to use a flat (very good) or lenticular disc (best) with this specific frame design to optimize aerodynamic advantage, which will be greater than the gains in rolling resistance associated with wider tires.
Again, the main points to take away are: 1) body position is the most important component in the aerodynamic TT equation; 2) one needs to think in terms of systems when selecting equipment (what are the gains in rolling resistance reduction offset by aerodynamic losses, or the opposite); 3) that frame and component aerodynamics are not appreciably impacted by sticking a body on the bike, no matter how lumpy that body might be; 4) just as one optimizes one's body position, on also needs to carefully select components that complement each other and which also complement a frame's/fork's design.