For our 'splendid night' JD2455945 we can have two estimates of the albedo of Earth:
One estimate is provided by the 'profile fitting' where albedo is directly determined as a model parameter of the Sun-Earth-Moon system, and another estimate is given by the B-V determination we have from the cancelling halos.
From profile fitting albedoB/albedoV: 1.217 +/- 0.003
From B-V: albedoB/albedoV: 1.202 +/- 0.005
The error on profile fitting results is propagated from the standard deviation of several determinations of albedo in B and V and thus is 'internal error'.
The error on the second estimate is taken from the paper and is 'internal' also - the dependency on error of the solar B-V value drops out.
The two estimates are similar but, given errors, formally different.
The fitting result is based on 6 fits in total - three in the B image and three in the V image.