But we don't dump boost between gears when doing a quarter, so does this really matter? Spool up only happens when we first drop the hammer.
(note, I don't know what I'm talking about here, so feel free to correct me... but I thought this was the idea behind NLS)
If you don't put enough load on an engine you can lose boost, too. Thenico is right in what he says about starting at 4th vs 5th, unless possibly we can spool infinitely and keep making power while doing it.
Gearing is all magic designed to deliver torque to the wheels. Remember that torque is torsional force. If the engine's crank turns over once and that corresponded with one turn in the tranny and that corresonded with one turn of the wheels, then your torque multiplier is 1 and a 100 ft/lbs engine will put 100 ft/lbs to the wheels. The shorter the gearing, the more torque you put to the wheels but the less time you have to make it before shifting. You can go 0-30 at 100 ft/lbs (assuming flat curve) and also go 0-3 at 1000 ft/lbs. Which is more practical?
I am now starting to think that it might be interesting to make a little program that determines acceleration given some engine output and gearing. This sounds like a classic optimization problem. Given a set of constraints, determine the ideal gearing required to maximize both acceleration rate and time required to go from 0-X.
Maybe we should put a CVT in the sonic.
And you will lose some boost even with NLS. In that moment of shifting the engine is putting out less exhaust (no load) so the turbo spins down slightly which will cause a hiccup in the boost level. I'd be curious how much mechanical forces buffer this effect.
If you are trying to make a 1/4 mile monster, your ideal goal should be to go through the traps right when you'd need to shift again in whatever gear you would top out at at those speeds.