Chris Perez's lens tests, for instance (http://www.hevanet.com/cperez/MF_testing.html), found almost nothing that could deliver more than 100 lpmm to the film; one very very good Xenotar, a couple of Mamiya lenses, and that's about it. The Tessar he tested peaked at 68 lp/mm (that includes the film's MTF, though).
That said, it's not totally clear to me if the diffraction limit is the sole driving factor for the sharpest aperture, or if there are other issues to consider---and most of the lenses, including the Tessar, in that test have their sharpest results at the f/8-f/11 you guessed, but at delivered resolutions far below that theoretical 136 lp/mm (or 187.5 lp/mm at f/8).
Is this confusion I have making any sense? It seems like, on the one hand, hitting the diffraction limit at f/8 or f/11 would require rather stellar performance from the lens, per the numbers above. On the other hand, the Tessar Perez tested (to pick one example; the other lenses do broadly similar things) starts to show dropoff at apertures above f/8-f/11, but it's doing it while delivering resolutions *far* below what the diffraction numbers would suggest---60-68 lp/mm in the case of that lens. You'd expect some loss due to convolution with the MTF of the film, and some due to in-camera issues like film flatness, but are those components really likely to explain all of the difference between 136 and 60 lp/mm, or are there other effects to be considered here?
I'm not trying to argue about anything, just trying to understand how to reconcile these different numbers I'm seeing. Thoughts?