This is a great question.
It takes energy to accelerate the short block (rotating assembly) of a motor. This is why a motor with lighter crank/rods/pistons will be faster than a motor with heavier parts...all else equal.
Everyone seems to agree that it takes more torque to produce a higher accelaration rate in a car and likewise, it takes more torque to provide a higher acceleration rate in the short block itself - before the drivetrain/car is ever introduced into the picture.
If you put a car on a chassis dyno with a 3.73 gear ratio, you will be able to plot the engine acceleration over time (like Al's example...rpm increase per second.) Let's say it's an average of 300 rpm's per second over the course of the pull. Now put the same car on the dyno with a 4.56 ratio and repeat. As you stated, it will be slightly different. Shouldn't be alot, but all else equal, the 4.56's will make slightly less torque at the tires. Why?
Plot the engine acceleration for the second pull (4.56's) the same way you did the first. It will be higher. Let's speculate a 350 rpm increase per second average. The engine accelerated faster during that second pull. And what did we say about accelerating the engine itself? It takes more torque to do it faster. Therefore, in the second dyno pull, more of the engine's total available torque at any given RPM was use to accelerate the short block rotating assembly itself and less was left over to accelerate the drivetrain.
Another example. When using an engine dyno, you can perform various different tests. Normally what you see on TV and stuff is a test where the engine is allowed by the water break to accelerate at a specified rate. Another popular but less known test is what is called a step test where the engine is actually held by the water break at stepped intervals - say every 200 RPMs. Guess which test almost always makes more torque/power? The step test. Why? Because during the step test, as the motor is being held and not allowed to accelerate, all of it's total available torque is applied against the break and none is being waisted in accelerating the bottom end assembly.
But we aren't talking huge differences here. The fact that a car may dyno a little less with a steeper gear ratio has no bearing on how it will run at the track.
One last thing along these lines. Think about how much a big block's rotating parts weigh in comparison to a small block and remember it takes torque to accelerate all those heavy parts!! (no offence to the big blockers here)
Sorry to deviate from the original subject matter....but she asked.
