Correct, there should be no shear stress on a spoke, (well except when that stick gets stuck in the wheel). Have you ever seen how rolled threads are made? they are actually deforming the metal into the shape of the threads, this is a stressor on the material as well and is stress hardening that section of the material, making it stronger (granted a bit more brittle as well). The theory still holds that the weak spot is the transition between the threads and the 'shoulder' when loaded in tension, there is no undercut or stress relief at that intersection to mitigate the stress riser, do the first thread concentrates the stress and thats where they are prone to failure (at the spoke end at least). The change in material properties from the stress hardening is also working against the you in this case since the two areas of the spoke will behave differently under load and deform at different rates.
keep in mind also that while rolled threads make the major diameter a hair bigger (im not sure if .2mm is correct, but lets run with it) it is still creating a minor diameter when it forces that material out, that minor diameter (lets call it -.2 assuming that the material that is moved all goes out) is still the weak spot in tension, the smallest area is where the failure if prone to happening (assuming no material defects are present)
keep in mind also that while rolled threads make the major diameter a hair bigger (im not sure if .2mm is correct, but lets run with it) it is still creating a minor diameter when it forces that material out, that minor diameter (lets call it -.2 assuming that the material that is moved all goes out) is still the weak spot in tension, the smallest area is where the failure if prone to happening (assuming no material defects are present)