Exactly. In order to get the maximum performance while climbing hills, you need to get your gearing just right, so that the engine is running at maximum torque while you're at the speed indicated.
The efficiencies I posted above were, more or less, maximum efficiencies (except for Belt CVTs. I had an email from a COMET engineer a couple of years ago that described the efficiency of a belt CVT and how it changes over time.) As gears wear, as chains stretch and bearings wear, and things 'settle' in to place, misalignments can creep into play. All of these things tend to reduce the total system efficiency.
Fabian, a 30 degree slope is tremendous - for every hundred feet horizontally, you rise almost 58 feet! It is a 58% slope... Most ranch-style homes have no more than a 4/10 pitch (40%), which is 21.8 degrees.
As far as the Nuvinci at 1:1 having increased efficiency - Probably, I would think. I believe the hemispherical surfaces where the ball bearings ride are not quite perfectly spherical, so that as the gear ratio increases, they are slightly closer together, which increases pressure in order to reduce slip. (Fallbrook mentions 'geometry changes' which lead to increased pressure to support increased torque transfer.) Increases in surface pressure to support higher torque transfer would also increase 'rolling friction' in the hub; conversely, decreases in surface pressure should also decrease rolling friction, and increase efficiency.
Also, there IS a thin film of fluid between the torque transfer balls and their mating surfaces. The fluid forms what the manufacturer calls 'liquid gears,' because it increases the shear forces (torque) needed to make the drive slip. Under pressure, the fluid acts to transfer shear forcees similar to elastic solids. There is also a very small amount of 'creep' or 'slip' inherent in the design. Ref Fallbrook's '
Traction Fluid FAQ'
It sure would be nice if someone with a Nuvinci was able to get it on a dynamometer, then, put a standard hub on the same bike/engine on the dynamometer, so we would have some real data to look at!
It wouldn't surprise me to find that the Nuvinci hub DOES lose 10% of the available power (or even more.) It makes up for it by allowing the 'motor' (whether that motor be electric, gas, or human) to run at max efficiency over a wide range of speeds.