Here are some more calculations:
Continuing with the same theoretical engines from my previous post:
To rotate one of the engines once, both pistons would have to cycle back-and-forth once. Let's find the total volume of air required to do that: 1.767 cubic inches (fills one chamber of a 4" stroke 3/4" bore piston) x 2 (motions per piston) x 2 (pistons) = 7.068 cubic inches. Multiply that by two again, and you get driving both sides of the drivetrain forward one turn: 14.136 cubic inches.
Remembering that that's in 60 psi, we can half it (this is an educated guess) to get the equivalent in 120 psi, our store pressure: 7.068 cubic inches. If we are using those fancy white plastic tanks (28 cubic inches), that's 3.962 turns of both sides of the drivetrain per tankful. We need some gearing.
Since 4 rotations to a tank is sort of ridiculuously low, lets see how this needs to compare the the 2010 KOP drivetrain (continuing the toughbox thing): In a conservative estimate, the CIM's will be moving at 2000 rpm. With an overall gear reduction of 8.693:1, that works out to the wheels on the robot turning at 230 rpm. With a 2.5 min match, that's 575 wheel rotations in a maximized scenario, 1150 for both wheels.
Now let's figure out how to fit 1150 wheel rotations into, say, 64 tanks. We know that one tank will give us 3.962 engine rotations on both sides per tank, so 64 tanks will give us 254 engine rotations. The gear ratio necessary for each side is easy now: 1:4.53.
Now, once again, back to torque: with the 2010 KOP gear reduction, the wheel torque of those two maxed-out CIM's on one toughbox will be 680 oz-in x 8.693 = 5911 oz-in. With one engine, we have 1600 oz-in. / 4.53 = 353 oz-in. - pretty weak.
You could get this to work out, eventually, but it would require an absolutely ludicrous amount of air tanks. With the 64 tanks in these calculations, you would have to make the robot rather light to get ideal preformance, not to mention that you would spend the entire morning of competition filling those tanks up to 120 psi for one match, at least with anything close to a KOP compressor.
Conclusion: it won't work
