Confused about gamma values for hybrid process

I could use some guidance on how/when to properly set gamma value when scanning B&W film, as I've managed to confuse myself a few different ways.

My goal here is to use digital scans solely as a "proofing" step for photos I eventually want to print using either VC silver gelatin, or alternative processes. So I don't care as much about "how to get the best digital image from this negative" so much as "how to get the best approximation of how the analog print of this neg will look with X process/paper" so I don't have to waste tons of time and paper doing trial-and-error printing.

Where I get flummoxed is that there are three different gamma values in play: the contrast index of the negative itself (defined by film/developer/time), the screen gamma (2.2, right?) and the exposure scale of the paper/process I eventually want to use to print.

What I've been doing so far is to scan to 2400dpi, 16-bit HDR RAW linear TIFF files in SilverFast 8, which seems to work as intended. The only adjustments I make in SilverFast are crop, rotate and flip. My understanding is that the HDR RAW format does not encode a specific gamma.

I then process the image in either GIMP or ImageMagick to take some "densitometer" readings of the raw scan before trimming both sides of the histogram, inverting, scaling down, and exporting to another file to visually proof.

But I'm never explicitly adjusting the gamma along the way, and I realize that I should probably be setting to 2.2 before manipulating the histogram in order to get the most "true" representation of the negative's actual contrast index on-screen... right?

And then, if I want to "see" some approximation of how the image will look when printed, should I apply the gamma adjustment for the paper/process-specific contrast after adjusting for the screen but before trimming Levels? Does it matter?

Or is this a fool's errand and I should just suck it up and start buying paper in bulk?

I have been printing on silver paper since 1958, as a kid. I wound up the last 15 years using Ilford MG IV with an Aristo cold light using the split filter technique, with great results.
Professionally, and for myself, I have been scanning for the last 25 years or so with many different kinds of scanners. The last 5 years I suspended darkroom work (but got into digital) for family reasons, but hope to get back into it soon. During this last 25 years I also learned a bit about Photoshop, so I have a good reference for both.
I retired a couple of years ago and am now, with a V850 Pro, using SF 8 software, beginning to scan negatives for much the same purpose as I think you are saying. I have always scanned 120 and 4x5 film as a "contact printing" option just because it was faster, even if only for selecting the best frame. I also played around with scanning a good final print, then the negative to see how easy or difficult it is to match them. The opposite ends of the dynamic range are not difficult to hit, remembering that the real range for a file is in the final output medium, but my attempt was to try to match the end points visually, so to speak, then see what happens with the rest of the curve. What you are asking about gamma, I think, gets to where I am here.
I was amazed to find out how different the low to upper areas differed between the two processes. (Between zones 2 and 7, so to speak). The placement and distribution of these intermediate areas was not easy to match without varying and inconsistent application of the curve and histogram controls within the scanning software, from neg to neg. Even then, more tonal editing may be necessary in PShop.
I am also noticing that just scanning old negs for sharing digitally with school mates and family, that my application of the curves and histogram in SF is amazingly varied from roll to roll, just to get a "normal" kind of image. Probably due to the range in films and developers over the decades, though most negs are well exposed and developed. I think I've been accommodating this variation with paper grade and exposure, without thinking about it.
My plan for the future is much like yours, if I understand what you are saying. Working the image up (exploring the possible final positive image) with the scanning and editing method, then going into the darkroom with a plan.
This would give me an image in my head which might speed up getting there in the darkroom, but l don't know, maybe save time with test stripping at least (split filtering takes two strips, one for each filter, etc.).
My take on using quantitative tracking of values in the curve is that it won't have value in the darkroom. Long way around for an answer, but it gives you my reference.