Roger.
I agree about box speed as a staring point, but here is what happened when I did so.
Being an engineer at daytime I thought "Let's do this acurrately once an for all. If I am lucky this process would help building the instincts needed when you are out in the field doing street photography with little or no time for fine measurements." The outcome of the story below is that the quality and the evenness of my negatives has increased dramatically.
Btw, I am not a zone fundamentalist. Whatever gives insight is studied carefully but later applied liberally.
The rules are there to be broken once you know them good enough.
Like when you are out in the streets with a Bessa R and a Leica lens and things move fast and you have to use (more or less) average-based measurements.
I took my 4 (5) favourite films. Tri-X, (TMax100), HP4, HP5, TMax3200.
Took an R8 with accurate shutter and exposed a large white sheet in a closed room with fixed lightning. Started at box speed. Ran one frame for each zone centered around zone 5. Developed in XTOL to standard recommendations and measured the frames with desitometer to record the curve. Ended up exactly where Barbaum said I would (in the youtube video).
Tweaked the process (exposure and development) systematically until I got enough info in the shadows and an intensity curve towards the highlights that could be easily scanned. Ran the frames of the final films through the scanner(s) and adjusted scanning and printer profile until the print matched the film densities using reflectomter to check the (glossy) inkjet paper.
This cost me all nights for a week, quite an amount of film, and it set me back 250USD for a used Barbieri densitometer/reflectometer but it was well worth it. It's amazing what you can buy cheaply today as the large labs shut down.
Findings were:
Tri-X: 160 ISO (320 ISO not too bad depending on the subject.)
TMax100: 100 ISO
HP5: 160 (or 320 ISO as above)
TMax3200: 1000 ISO.
Had to take 10 to 20% off development time for different cases for the highlights to remain printable.
All data are of course personal in the sense that other people with different details in the development process may find different results.
This has all been done before and you can find the procedure it in textbooks. The point is to get to know your own processes. - And then go and have fun.