SinoCORFinder does not `fliplr` the second sinogram half.

Looking at the SinoCORFinder code, it seems that the second half sinogram is not flipped, contrary to the second radio in the CORFinder equivalent.

I can fix this in the "big" MR that I am working on.