In my understanding much of the sensitivity of coherent Raman techniques is due to the preparation of vibrational coherency in a volume of the sample. When the number of centres is reduced, this advantage decreases and I wonder if the final result in terms of sensitivity is similar to spontaneous Raman in the case of a single molecule. Of course there may be other reasons to measure single-molecule vibrational spectra using coherent techniques, but, as far as ultra-sensitive detection is concerned I cannot see the point. Can anyone explain what I am missing?