Thoughts on further experimentation tonight...
First of all the rig - K-1, DA35 Macro, and the usual 49mm UV filter on the front (because most of my work shooting is in a wet environment with toxic chemicals that I want to keep off the front element), which offers my Asahi Pentax slide copier something to grip on. With the copier mounted directly on, the lens sees the frame of the copier and the entire slide within it. The vignetting, if any, at this distance is clear of the actual image I want to capture (bonus!). The illumination is from a bright 50W-equivalent LED downlight in the ceiling.
In-camera settings (fixed throughout) - ISO 1600-3200 to keep the shutter speed up for these impromptu tests, monochrome filter, colour inversion filter, f/8.0 for reasonable DOF on the film plane, Av mode, downsampling set to 2.0MP because I'm really only fooling around with exposure and don't care much for detail at the moment. From time to time I went to crop mode to play around with framing, but crop mode costs me the top 5% of my picture and still needs the bottom 5% of dead space cropped away; full frame leaves a bright frame around the actual film image. If I'm going to straighten and crop, I'd rather do it with one setting in post.
The variables - third dial assigned to EV compensation.
The images - randomly selected from the past year or so of B&W film processing (yes, I've been doing it for that long!).
What I discovered: Regardless of what film or camera I had used or what development, +2.0EV got me a pretty good insta-result MOST times, regardless of which strip I pulled out of the negative folder. These are the sorts of results I could put straight on social media with a bit of frame-cropping and straightening and they would be just fine.
Sometimes +2.0 wasn't quite right, and I had to go either higher or lower to get an optimum result. But this is really no different from making a contact sheet and playing around with the exposures on that.
With the scans from these, even without frame cropping, I can have a quick look at all the pictures as positives (black and white photograph equivalents) as soon as the film is dry enough for scanning, cluck over the best ones, and from there rig everything up very, very carefully to take full-scale RAW shots from the negative of the ones I might want to make physical prints from or give to people as high-resolution JPEGs. Then I can take those as negatives into Raw Therapee with the exposure compensation factor already applied and do the monochrome conversion and negative inversion under the control of a significantly more powerful processor than the one in my camera.
That's the idea anyway.
The key lies in thinking like a film photographer rather than a digital one. In this case my sensor and processing system are the "paper" via which the negative image is inverted back to a positive. Of course if I'm taking an image for keeps then the whole business will be on a tripod and the lamp will be parked in front of it, and I'll be using live view and a cable release at the lowest ISO possible, but the principle seems well and truly sorted out now. Raw Therapee is by no means the ideal JPEG manipulation tool, but if all I am asking it to do is the same cropping and rotation (something it's pretty good at), there shouldn't be much harm in that, especially since the "raw file" in question is an actual, physical FILM NEGATIVE.