Scott...with the dynojet the torqueload on the motor is the weight of the drums, which as you say is unchanged regardless of the cars gears. However...put the 4.10 in the rear in place of the 3.42 and it will spin up through the engine rpms faster due to mechanical advantage. So with 4.10s you would see that the car goes from say 3000 rpm to 6000 rpm in say 8 seconds. Which would be faster than the say 8.5 spin up with 3.42s. However...look at the dynojet readouts for mph of the drums. You would see that the the time to accelerate from X mph to Y mph will be less for the 3.42 gears

So you would see what looks like two conflicting stories until you seperate them.

Time to accelerate the 3.42s from 3000-6000 rpm= 8.5 seconds
Time to accelerate the 4.10s from 3000-6000 rpm= 8.0 seconds

it looks as though the 4.10s are better..untill you compare that to something that will better show the work done at the wheels. It may look something like this

Same dyno runs as above...just reading time to MPH instead

with 3.42s in 8.5 seconds went from 40-140 mph
with 4.10s in 8.0 seconds went from 38-119 mph

if you did the math you would see that with 3.42s your accelerating at an avg speed of 11.76 mph per sec
and with 4.10s it would be 10.1 mph per sec

More work done at the wheels with 3.42s, means more tq/ hp applied to the rollers of the dynojet.

This could easily be tested on Johns dyno. But I believe its right.