On "the other forum" I believe the number tested with a sound card O-scope was something like 6 seconds. And that was attributed to the engine acceleration over idle.
This figure goes down if you use that energy stored in your "spun up" engine. So if you're engaging the clutch before the engine falls back down to idle - that 6 second figure goes down.
I'll stand by skewbe's testing methods - unless someone brings
empirical data to the table. Anything less is speculation.
In any case - a good hypermiler has the engine cut before reaching the stop light for that little extra something
