Based on what I've read...
Assuming taller gearing means the engine is spinning slower at some given speed. In order to make the same amount of power at that given speed, with the engine spinning slower, more fuel, and more importantly, more air must be in the cylinder, which reduces the difference in pressure between the volume of the expanded cylinder and the crankcase durig the intake stroke, which results in less force needed to turn the crank. These are pumping losses, and are a problem to some extent with every spark ignition to date. Lower engine speed also results in fewer friction losses since the engine turns less per a given power output. High EGR rates as well as forced induction also reduce pumping losses.
Driving the BMW at 75 mph in fourth gear yielded 26.4 mpg. In sixth gear at the same speed, it got 29.7 mpg, a 12.5-percent improvement. In lower gears, the engine simply spins faster, generating more friction and sucking harder against a partly closed throttle.
I think the spinning refers to friction losses and sucking refers to pumping losses. The way to test this is to drive a nice round trip with no stops at the same speed in two different gears... If going from 4th to 5th increases economy, then going from 5th to a taller 5th should have the same effect depending on how much taller the new gear is. GM is also planning a new series of six speed automatics, with taller OD gears.
Originally Posted by Green Car Congress
The company recently introduced three new members of the six-speed family. Most of GM’s six-speeds feature a wide, 6.04:1 overall ratio compared to 4.0 of typical four-speed automatics.
Originally Posted by FormulaTwo
I think if i could get that type of FE i would have no problem driving a dildo shaped car.
Yes more fuel per stroke but less strokes per second so you can come out ahead a little if you are on the right point on the thermal efficiency curve.
One thing factors into the rpm that I didn't appreciate which explains a lot - the Synlube site has a friction vs speed graph of metal surfaces rubbing with oil vs synlube and you all should look at it. It really explains a lot about what is happening when we lug and over rev the engine.
Some anecdotal evidence would be what Honda did. One the high FE 92-95 versions of the HB (CX and VX) they dropped the final drive to 3.25 from 4.06 on the DX. They also reduced weight and CD and used different engines, but if shorter gearing was better for FE, why did Honda do the opposite? The EPA figures show that the CX and VX FE is clearly better than the DX.
Since Honda engineers are much smarter than I am, I will let thier actions speak for me.