Apparently this question was edited while I was answering it. To the question, "Can you generate lift?": in a word, no. For two main reasons.
First, lift as we know it, and as codified in the classical lift equation, is generated in a collisional gas, i.e. the average distance between molecular collisions ("mean free path") is much smaller than the dimensions of the object generating the lift. This is not true at orbital altitudes. (See the discussion in David Hammen's answer to "Why are LEO satellites not aerodynamically shaped?") Due to the absence of molecular collisions the molecules traveling "over" the "wing" cannot have their paths deviated to follow the contour of the wing as the traditional Navier-Stokes approach to the flow would suggest. So the classical lift equation can't be used.
With the classical lift mechanism invalidated, the only other potential source of lift is molecular deflection, where incoming molecules collide with the surface of the object and reflect ("bounce off") that surface in a quasi-specular manner, resulting in a net momentum exchange that produces a force. But to get that force you need the quasi-specular reflection, at least to some degree, and this is tied to the "Accommodation Coefficient" that has been in the theoretical and laboratory literature since the 1930's, and in tests in space since the 1960's. Work by Kenneth and Mildred Moe indicate that measurements of the Accommodation Coefficient are consistent with entirely diffuse and low-velocity post-collision molecular trajectories, suggesting a large fraction of the molecules adsorb onto the surface and then are re-emitted: no specular reflection, so no deflection force.
That said, you can get some force component from the re-emission. But the re-emission speeds are at the thermal speeds for the temperature of the surface, and those are much slower than orbital speed. So the force derived from the re-emission is much less than the drag force, and L/D is so small it is worthless.
By the way, the research on atmospheric drag in orbit quoted above indicates that in very low Earth orbit there is some variation in the drag coefficient (and thus the Accommodation Coefficient) with vehicle shape, but above that it appears to be independent of vehicle shape. So it wouldn't matter much whether the solar sails were thin-film or rigid.