M Sport said:
Cervelo and Sky tested the theory. For sprints out of saddle sprinting was superior. For track it's in saddle.
I don't think it's as simple black/white as that, as track geometry and rider morphology and performance variance standing to seated may change the equation. For some it might be a function of getting into the saddle earlier, rather than staying in saddle the whole time.
The issue is using the data available in unique ways to answer the question, and work out if there are other ways that lead to a speed improvement. It might be that a rider is technically uncomfortable with the transition at certain parts of the track and it needs practice before an overall improvement can be attained. But such improvement should be detectable in one or two sessions, and not require months.
Sprints efforts that are not TTs and involve more than one rider also mean the craft of sprinting and track/road position matters, a lot. Which is why slower riders in the flying TT still win sprints.