First of all, you still have unprocessed data from AF?!
I would have thought you'd be best using the same stretching method for both L and RGB? If using a masked stretch (MS) for one, use for both. If using curves (HT) for one, use for both. Just that I've ended up with funky results (not in a good way) if the L and RGB are stretched (at least that initial big stretch) differently, esp in stars.
Re the donuts in the core of bright stars from MS, most recent images of mine have been HDR just for the star cores to alleviate that. Just at end of session take some shorter subs of LRGB. Don't need as many as the main/long subs, because you're only using them for the brighter stars, less worried about S/N (maybe 5 subs, whatever min to get BPP to stack). Not that helpful for your data from AF, but maybe for next time you're out?