I like your train of thought. You absolutely want the right amount of taper. Regarding wood strength being proportional to area: it can't be as simple as that

I like your train of thought. You absolutely want the right amount of taper. Regarding wood strength being proportional to area: it can't be as simple as that - wood is a strongly anisotropic material, so direction of loading becomes critical to understanding what's going on. Natural trees have defects, and they "know" it. Limb attachment is messy business - in many species they are frequently defective. Doesn't reduce well to modeling actual limbs as simple beams or whatnot. Starting to understand thigmomorphogenesis and how trees deal with it in terms of load shedding, compartmentalization, and reaction wood growth was one of my favorite "ahhh ha" moments.