Oh, I wouldn't go so far as to call the 35 mm Planar "fairly poor wide open". The MTF values are not as stellar as those of the rest of the Contax G lens range (and perhaps that's all you meant), but they're still similar to or better than other good 35 mm lenses from Leica (pre-ASPH), Canon, Nikon, etc. I've noticed that fast 35 mm lenses from many manufacturers have historically displayed poorer optical performance than their slightly wider or longer siblings. If I had to guess why that is so I'd blame:
- The tendency for fast 35 mm rangefinder lenses to be based on a modified double-Gauss design, which is already uncomfortably stretched in 45-50 mm f/2 designs. Lenses wider than 35 mm require different, more expensive designs (often retrofocus), which are better tuned to their focal length and aperture, so perform a bit better at much greater cost. Longer lenses are well-served by the double-Gauss design.
- The curvature of field problem that results from trying to use an almost symmetrical design (truly symmetrical designs suffer from poor flatness of field because any curvature created by the first half of the lens is doubled by the second half).
The Contax G 35 mm is a double-Gauss design with a thin positive meniscus inserted between the two halves, probably to flatten the field. This design is appealing for 35 mm lenses because it's small and light despite the specification, it's affordable to compute and manufacture, and any slight compromises in image quality have to considered in the context of how a fast 35 mm lens is typically used: often handheld at slow shutter speeds with close subjects near the centre of the frame. The 35 mm Planar is the easiest lens in the G range to shoot handheld in low light, and in those circumstances it should perform admirably based on the MTF curves. Note also that it's among the lightest and smallest of Contax G lenses, and also quite affordable, justifying the use of the modified double-Gauss design. I can't vouch for its flare and ghosting resistance though, which are also very important in low light.
Mike, you've probably figured this out by now, but 10 cycles/mm is roughly equal to 10 line pairs/mm (not 20). The difference is that cycles follow a sine-wave pattern while line pairs form a square wave. The difference isn't very important in this case.