Siril is a very powerful and flexible tool, but I don't think there is any one right way of doing things. I thought I would share my current workflow steps here for mono processing to see how they differ from other folks.

Pre-processing
I have a python notebook with some code to pull new data from a usb stick and place it in a project directory tree (i.e., Root/Target/Filter)
It also outputs a summary of how much time I have on each channel of each project so I know when I'm done capturing data (for now)

Per channel stacking
Add all data from a single filter in the UI. Name prefix: <project>_<exposure time>s_<filter>
Apply calibration data (bias, darks, flats)
Select a reasonable reference frame (not blurry or cloudy, object framed as intended)
On the registration tab, check number of stars detected, adjust threshold so the number is between 100-1000
Register using global alignment method, check 2x drizzle box
Sort list by FWHM, review plot to estimate number of outliers, either deselect manually or select FWHM weighting and select % of frames to use (usually 80-90%)
Stack with "average stacking with rejection", normalize with "additive with scaling", reject with "sigma clipping" (usually 2.8 high/low)
Save output to a seperate folder: <project>_<exposure time>s_<filter><frame count>

Cross Channel Alignment
Reload all frames as a new sequence, include channel order in the sequence name (e.g., BGRHOS)
On the registration tab, recheck number of star in each frame (number often goes way up after stacking)
Register using global alignment method, uncheck 2x drizzle box
(optional) Perform background extraction for each channel, 8-16 square grid + some manual point adjustments, 2 degrees of freedom
Auto-stretch each channel, save as 16-bit tif

RGB Specific
Use RGB compositing to combine RGB into a single color image
Photometric color calibration
Auto-stretch, save as 16-bit tif

Outside Siril
Starnet++ v2 for each channel
Load all layers into Photoshop
Subtract starless layer from RGB image, set as top layer, change type to 'Screen'
Use camera raw filter to defringe, increase saturation, adjust star size
Adjust each starless layer with some combination of 'Shadows/Highlights', Camera Raw Filter, Topaz Denoise (with sharpness=0 and recover original=40-95)
Use spot healing brush on artifacts (usually from Starnet)
Set each starless layer to 'Linear Dodge (Add)'
Apply a gradient map to each narrowband layer (check apply only to layer below), set the low side to black and the high side to the desired channel color (e.g., SHO, HSO, HOO, or somewhere in between.
Final adjustments as needed (cropping, rotation, overall brightness, etc.)
Export as, save as jpeg with max quality, but resolution slightly smaller than original.
Like