I see now. I think my approach is more consistent with the standard model, at least in the sense I am concerned with comparisons. Ultimately for me the speed point is a means to an end, not a target in and of itself, the end being maximized local contrast. If a film/processing combination delivered a perfectly straight line from Fb+f onward (ie there existed no "toe"), my speed point would be nothing more than 0.01 above Fb+f.