As I said, I think more distance between stops would increase the benefit of this. But I'm not sure. I stopped every 3/4s of a mile or so. Anyway, leaving the engine on would yield the same results except mehbe slightly closer, there's really no reason to leave the engine on though,
Valid point Zugy. Perhaps a chain is overkill though - something (e.g. wood) of varying thicknesses placed beneath the gas pedal would be easier if you wanted to do the test using a physical restriction.
SG users could also watch TPS and/or LOD values.
Then again, knowing throttle opening values exactly may not help other cars since they have different fuel / throttle position maps.
Maybe timing the slow vs moderate rates of acceleration would be a better comparison point.