Actually one can not do a film speed test on other than
zone 1. Zone one's 0.1 density establishes the film speed.
Two variables are involved, exposure and development.
In practice one adjusts exposure while holding
development constant.
I understand that increasing development time will increase contrast. I suppose this plays havoc with trying to establish a "film speed" based on a zone 5 density since increasing development will move the zone 5 density up. IF you follow traditional thinking about this, but forget that for a second.
I'm trying to reconcile incident metering, which meters mid-tones, with an idea that negative film's speed is based on the shadows and development can be varied, so the midtones move around on the negative! As you say, it only makes sense to meter the shadows in this case, but the point of incident metering is you just meter the light and let things fall where they may.
The zone system is great because you meter the shadows and figure out exactly how much shadow detail you need and expose just that amount. But with an incident meter you wave it around and get your incident reading and let things fall where they may. But it seems to me that the exposure recommendation is too low if you are just relying on latitude to have adequate shadow detail for the most part. In other words, if you have to subjectively correct the reading the incident meter gives you, depending on how much shadow detail you think you need, well now you are guessing as to the proper exposure again and you might as well be using a spot meter.
If I'm taking a more or less naive incident-light meter reading and letting tones 'fall where they may', why should I not put the meter reading such that my middle grey exposure ends up right into the middle of the curve...with equal latitude below and above...when the negative is developed 'normally' for a normal contrast index? Sure, you'd be exposing 'more than necessary' in flat light, and pay a grain penalty compared to metering off the shadows.
Say you establish a development time so you get .65 contrast or whatever works best, then the zone 5 density won't change, and you can set the film speed right in the middle of the curve. I mean this is what you do when you shoot transparency film...the process is standardized. The development does not change so you just meter the middle and let things fall on the film where they may. How would I go about establishing a development time and zone 5 negative density if I wanted to do this with negative film?