The ranking of filter combinations does not change whether a window
from 1 A to 0.5 A used for the claculation of selectivity, except
for the SHG.
NB : the assumption made on the transmission profile of the SHG
(Lorentzian curve) is very conservative. Actual profile is steeper.
The selectivity of the mica-spaced 0.6 A Ca K etalon is lower than the
selectivity of the double-stack 0.13 nm because of the large wings of
transmission of the mica-spaced etalon:
Tolerances on the CWL for double-stack combinations based on Alluxa Ca K 0.1 A filters
The question is "If the CWL filter is not exactly on Ca K, what is the impact on the selectivity of the combination ?"
In the following table, the 1rst column shows
different assumptions regarding the difference between the CWL of
the filter and the center of Ca K, while the second shows the resulting
selectivity.

We can conclude that we can double-stack two filters with a difference
of CWL of 0.3 A. The impact on the selectivity (compared to a
combination with the same CWL) is negligeable.