As you said in your PM, Barney, test shots are probably needed to work out a usable function which would interpolate the desired values to use.
I'm thinking of taking several (many) shots at different settings - different isos, apertures, shutter speeds etc - noting which shots have good exposure and work out a function from them. It would be an interesting exercise anyway just to see whether the camera operates the same at settings it wasn't designed for. Like whether there are discontinuities in the function at some point.
