My understanding of weak gravitational lensing is that it assumes random alignment distribution of galaxies in order to estimate statistical shear and convergences, which are used to estimate matter distributions. This might be subject to noise that depends on the granularity and presence of galaxies, and the expected ellipticity of the galaxies.
matter sources will deflect light in several wavelengths in different amounts. Also, there are different backgrounds at different wavelengths that should be deflected consistently by real matter sources, specially dark matter.
Is such kind of multi-spectral, multi-background correlation analysis a standard procedure to validate dark matter concentrations? I'm particularly interested in the case of the Bullet Cluster, which is kind of the pivotal example to support dark matter.
For instance, CMB light should be deflected by dark matter just like far-away galaxies. I'm guessing without much knowledge about the subject that convergence of CMB is hard to estimate because it is mostly undistinguishable from thermal fluctuations (possibly with the exception of polarization of the CMB, which I won't discuss). But if we could correlate lensing/fluctuations of the CMB with galactic lensing (which is at different wavelengths, and at a different background), then one could in principle tell what features of the CMB are due to fluctuations and which one are due to actual lensing effects. Only real dark matter concentrations will persist across the spectrum and across the background. Measurement noise should be substantially reduced if done properly.
Does this analysis sounds accurate?