They aren't "made-up." They're based on actual wind tunnel tests by a 3rd party.
http://velonews.competitor.com/2012/...visited_256023
http://velonews.competitor.com/2012/...urchase_264284
And yet, the Cervelo S5 whitepaper you linked demonstrates that, in fact, aerodynamic benefits increase dramatically at higher speeds.
At 25kph, the S5 saves you about 5 watts. At 40kph, they claim it saves 25 watts (only 5w off the Velonews numbers). I'm going to go out on a limb and say that 25 watts > 5 watts. (Page 11.)
No, it really doesn't.
Wind tunnels test different yaws, to make a reasonable approximation of real-world conditions. No one would ever say that what you encounter in a wind tunnel is
exactly what you'd see in the real world. Nor did I say that wind tunnel numbers are useless -- which is why I cited them, with the proverbial grain of salt.
I'll stick with Velonews, kthx. Their numbers aren't necessarily perfect, but at least they aren't an interested party.
We should also keep in mind that in real-world conditions, we don't always see these theoretical advantages bear fruit. My favorite example for this is Mark Cavendish, who spent years sprinting on non-aero bikes (e.g. Scott Addict, Specialized Tarmac, Pinarello LSD Meltathon or whatever it was...) beating riders on aero bikes. This year he's riding a sprint-specific aero bike (the Venge), and getting pipped at the line by, uh... riders on non-aero bikes.
It's difficult to draw a specific conclusion from this. It could be due to problems with the lead-out; Cav may have had a bad line one day and not enough recovery from climbing on the next; this year's crop of competitors could just be faster; and so forth. But it may also be the case that saving 40 watts when you're cranking out 1600 just doesn't wind up having the kind of impact we'd expect, or that a harsher bike fatigues the muscles and somehow slows the sprinter down at a critical moment. We may never know.