I raised a question a while ago regarding weak gravitational lensing of galaxies and the CMB. With all the fuzz with the BICEP2 data, I think it is time to raise even more questions about this amazing subject.
One of the assumptions being applied is that if we remove all the gravitational lensing happening to the CMBR, then we are only left with the "non-lensing" CMBR, and any B-mode polarization left over must be primordial gravitational waves
What is not clear is why can't be assumed that all the the remaining B-mode polarization is due to weak gravitational lensing of unseen dark matter. Note: Remember that, given the question I referenced at the beginning, we use gravitational lensing already to 'measure' dark matter (well, at least extra-galactic dark matter distributions)