Well put Dan. I think we are having what is substantially a symantic argument.
"But where are at least 2 of the fingers on the string and the weight of the bow hand at full draw and release?"
Two fingers are still above dead center of the string, and the force on the bow hand is below dead center of the bow. This is the inherent asymetry positive tiller combats.
"At or below center on the string and bow respectively."
No, above in the case of the string, although granted not much above, and in the case of the bow, yes, below dimensional center, in the typical scenario.
"My argument is that the forces applied to the string and bow are below center making the lower limb effectively shorter during draw."
The lower limb, and the lower string segment, cannot change their lenght during the draw. The fulcrum can change, by healing the bow, and because the bow (if heavily positive tiller) tends to rotate upper forward during the draw. While admittedly difficult to mitigate entirely, this is not actually desirable, imho. And it doesn't really address the issue of stress, or more correctly strain, with regard to the relative strain on the upper versus lower, unless you are suggesting the lower should bend more than the upper at full draw.
I think we strive to bave the forces balance, but tend to err toward the side of overworking the upper limb, by having it be both shorter and bend farther. The differences in practice are actually quite small, getting down into the less than 1" area (lay your three draw fingers on a ruler, nock at 4" above with 3" of finger width, or 3" - 2 1/2" are typical), but again typically erring in the area of 1" or less ABOVE dead center.
As regards draw force on the lower limb, it introduces a very interesting aspect of the equation, bow rotation. If by "more draw force" you mean stiffer lower limb, the bow MUST rotate in the hand. It's simple teeter-totter physics and does happen in practice. Don't know how you suspend your bows on the tiller tree, but I use a single point of contact, and usually not a very pronounced throat, to capture the bow. Bow rotation is easy to see when I work a bow on the tree. I would argue we want the bow to rotate upper limb forward, if at all, versus the opposite, in order to "steer" the arrow away from the shelf on the loose.
More to the point, don't we want to minimize rotation where practical, minimize the inherent imbalance a shortened, weakened upper limb represents? This is the area that I find most interesting, where the potential benefits hide. When I investigate folk's understanding and application what I find is not much consideration of the alternatives, and a good deal of discontent with having been confronted with the question. As a mind experiment, what would the effect of having the arrow pass, simply as a proxy for positioning the nock point, be at or nearer the dimensional center? It's essentially just a little step forward from 1 1/4" above versus 2" above. What about 1/2" above, or dead center? What's the downside? What's the upside?
I just think the topic merits more exhaustive treatment than it typically gets, and that it sure as hell enjoys a well earned reputation for repetition of misinformation at the most callous characterization, or miscommunication at the more generous.
I sincerely appretation your participation, and again, I apologize if I've stepped on your toes.