Could we use gravity lensing to directly measure the difference in redshift of the same star at different points in time?