Flat problems

I have been having trouble with flats for the OSC and RGB filters on Pier 5. They are not correcting out the dust donuts.

Example attached (from Job 2429). Processed in WBPP and STF boost stretched. No other processing. It does not matter whether I generate the flat masters from the raw subs in WBPP or use a previously generated master, the donuts are stilll there. They are not so obvious until you remove the stars for further processing - then they are obvious.

It does not seem to affect the HaO3 and S2O3 filters; at least to the same extent.

Do other users see the same?

Do we need new flats (I notice that the current ones are from July)?

Is there a way of removing them that people have found reliable (I am trying to avoid GAMES as it is a pain to process out the dust donuts)?