Titanflexr wrote:
To use these data, it seems like we need to take the ambient wind data measured on a "representative" course (people can debate which course it should be, but based on the data in the article there seems to be a low sensitivity....so go with Kona since everyone puts that on a pedestal). With that wind data, you could take the known drag of different bikes/wheels at yaw, plug in a given rider wattage, and do a simulation run to see what the time savings is. The output will be different for a "pro" (300W) vs. a "solid AG" (200W) vs. a "bucket-lister" (150W). If you know your watts (or can backwork from you bike splits) you can see what setup works best.
The time at each yaw angle on the graphs is a bit deceptive since climbing at low speeds leads to higher yaws, but on a part of the course where aero gains are minimal.
I was thinking along similar lines, but was thinking you could have a "yaw distribution" inputs to Best Bike Split, not just a single CdA number.
But this would require some sort of standardized representation of such a profile. Which I don't think exists right now.