On Kerrison. The comes from outside the sport.
They can confuse the 0.0005% gain, as the aggregation of said gain, not appreciating that they are conflating the trees for the forest.
The Australian track team got sidelined by Braislford and Keen MO. Taking their team pursuit into a wind tunnel to determine the order.
Now, I am a track neophyte, I have zero appreciation with the on the boards TP strategy, and their order. I do know that circa 2000, Bannan et al, thought Graeme de Bruijn was the best starter in the world.
My point is, for the potential 0.0001% from the aero coefficient by changing their order, they are now susceptible and sacrificing their optimum strategy in their order. And they can't perfectly model the aero coefficient when there could be another 4 100 metres in front and there is a lead rider swinging off with his unique technique singing up the bank every half or whole, or 1.5 lap.
This aggregation of marginal gains is bunkum. Its is forest from the trees stuff.
I think D-Queued made a cycling aphorism of acuity, that (words to effect of) "marginal gains aint the rounding error relative to a comprehensive doping program of Ferrari proportion".
These, "marginal gains" are too marginal. Best to get the major inputs dialled to 100%. Doping foremost.
Heck, even a custom frame in Reynolds Columbus or Dedaccai tubing on the ideal geometry would be an improvement over the stock Specialized carbon models out of a Tapei tech engineering workshop. That would be a gain. but few teams have ever had the money to be able to afford that opportunity cost eh. USPS? Sky? Mapei?
NB. The Aus Olympic cycling managers would have done better putting the money into getting Bobridge and Hepburn some alcohol education