Wouldn't rim weights also dramatically affect this test? The 30mm Velocity Deep-V is 520 grams and I would have to expect Topolino's 30mm deep section rim to be a bit heavier since they have 12 fewer spoke hole drillings, with the '07 Rolf Prima Vigor being the heaviest yet owing to it's 34mm deep profile and having only 20 spoke hole drillings. Conversely, the MAVIC Open Pro rims are around 470 grams and the Campy Eurus rims are likely somewhere around 500 grams. Weight should have some influence, but once again there are so many other uncontrolled variables that the inertia factor simply gets blurred and you look once again to the overall results: very tight and similar profiles where the most obvious difference -- the number of spokes -- seems to be the only big discriminator in this one experiment that clearly favored lower spoke count relative to having a slower rate of speed decay.

does anyone know exactly what percent of the bike/rider combination this wind-cheating technology actually represents?

As always, there's a plethora of data on Analytical Cycling

Cervelo has a pretty good page on this which appears to be identical to this page.

And, well, if you have a lot of time on your hands: http://www.sheldonbrown.com/rinard/aero/index.htm