I have filters that are sold as f/3 filters and they work fine at f/5.
Now I may be wrong, but when other companies sell filters with a speed, that usually means that it will work at slower speeds, but past the "rated" speed, the bandwidth shift will cause the outer part of the field to be at a different band than the center.
If the filter only worked at f/2, then the rays entering from the center of the cone (shallower angle) would be diminished so you would not be in band at the center of the field which is typically where you want to place your subject.
Maybe I am wrong though, but my filters rated for f/3 works perfectly well at slower speeds.
I guess that's because the bandpass of your filter is broad enough and it's actually calculated to just work at f/3 (may be not 100%) and to be optimal in slower scopes.
Baader says very explicitly that these filters are only for f/1.8 to f/3.5. They don't publish the spectrum because it would look awful, although it works for the actual purpose.
In a former thread here at CN there were some spectrum posted from regular filters measured tilting the filter some degrees. The change was obvious. A simple test with your eyes also shows roughly the effect of changing incidence angle.
Edited by sanbai, 30 June 2020 - 10:41 PM.