In other words situations where we see twin images of a very distant galaxy because light curves around intervening mass, without sufficient mass observable to explain the deflection?
I was under the impression that we could map dark matter localized to a precision that tended to rule out measurement error or a general divergence from Einstein's laws, unless it had a peculiar non-uniform distribution.
Edit: maybe somebody can explain these maps - http://www.sciencemag.org/news/2017/03/scientists-unveil-mos...