This is a much better than my original take on the Flaming Star from brighter skies and entirely in broadband. This has the "quadfecta" of goodness compared to that earlier effort: more time on target, a narrowband assistant, darker skies, and more resolution. The only downside is a smaller field of view.

I think it was a worthwhile tradeoff though.

Here's how it was processed:

H:
dynamic crop
deconvolution
generalized hyperbolic stretch
starnet2
tgv denoise
LHE (2 scales)
curves for contrast

Each R, G and B:
dynamic crop
DBE

RGB:
channel combination (RGB)
PCC
generalized hyperbolic stretch
starnet2 (saving stars)
tgv denoise
LHE (2 scales)
curves for contrast
NRGBCombination in H (this makes a new image)
pixel math in the NRGBCombo using a lum mask
MLT sharpening
pixel math in stars
EZ Star Reduction
slight dehaze in photoshop

stars:
very slight MLT sharpening

I ended up doing some different things here from what I anticipated when I started. I thought I would replace the luminance in the RGB with the H but that effectively destroyed the reflection component. Next I tried NRGBCombination and that worked but even with small scale values for the H was clipping some red highlights. So I next tried doing some pixel math combinations of H and the luminance but while avoiding the clipping they tended to lessen the reflection tendrils too much. Finally, I hit on the idea of using a lum mask on the luminance and then overwriting the H onto it. That more context aware blending really seemed to keep the best of both worlds...except for some glare from AE Aurigae. That was tamed, at least a bit, in photoshop.

I also did some sharpening on the stars, something I don't normally do, but without it they seemed just a bit too fuzzy. 

I actually had L, S and O data that I didn't use. The reason for not using O was simple...there wasn't any signal. The S was there but so weak as to not really be helpful. I had planned to use the L but for some reason I could not add it to the RGB in such a way that it actually helped. Plus the starnet artifacts were much worse in the L image. Ultimately I decided to skip it as the image was better without it. That probably says more about me than the dat though.

Still, despite its challenges, this was a fun image to work on. It's a fascinating region and always worth spending time there!