I follow all that. It's only about clamping force, and torque is how you measure it. If you have to know, get data on measured and "true" torque with varying length extensions, and apply multiple linear regression or multivariate analysis to see which formula really fits. I'm sure there's a learned engineering paper out there on setting recommended torque ranges to accomodate the widest range of wrench error using statistical analysis of true vs indicated.

I'm going to do it just like always, 95% of the bolts I tighten I do by touch, stuff I want to feel good about I torque wrench (mainly heads, mains, and rods), based on learned bolt break points mostly, I guess.