In Reply To:
How then does one account for the following (from the article)?
"In general terms, the total drag of a cyclist will consist of 80% tire rolling resistance and 20% wind resistance at 10 km/h or 6 mph. At 40 km/h or 25 mph the numbers will reverse, with total drag consisting of 80% wind resistance and 20% tire rolling resistance."
&
"When pumped up very hard in excess of 9.5 bar (~140 psi), rolling perfomance will improve quite dramatically."
It seems to me that these two points taken together mean that going fast would require a tire which, together with the rim formed as smooth an aero section as possible, was quite wide, and was pumped up very hard indeed.
The higher the internal pressure the less the tire will deform when loaded, which I would have thought would give a contact patch more in line with the fatter tire diagram.
Less deformation is not always what you want. It works on a track (or steel drum), but can slow a tire down on asphalt roads -- even smooth ones. The tire bounces over imperfections, rather than deforms around them. The net is a loss of speed. A tire that is supple, and deforms readily, and returns to shape readily, is what makes a low Crr number. So, hard is not better, again except on tracks and metal drums. The same feature that makes a fast tire on a test rig (one that deforms at the contact patch and returns to shape without heating up or squiggling on the glue) makes it fast on real roads, but only so long as it isn't pumped up too high.
On the other note: The lower one's CdA, the more that Crr matters. It eats up a greater proportion of the power. A rider with a CdA of 0.32 doesn't have the Crr worries of a guy with a CdA of 0.25.
As always, the idea is a nice slick aero setup, with low Crr tires. Folks can argue the numbers all they want, but I have yet to have a triathlete (about my size) step forward who can go as fast as guys like Jens and me at the same power output.