Hoski :- I understand your comment about the additional interface but as the thread is a very fine spiral and the shim is trapped between the cap and the top race and or spacer and then the whole assembly is clamped with the steerer-stem clamp - I cannot see anything moving or be compromised.
I guess my dislike comes from shimming something that simply doesn't need shimming.
I would agree with
@andrew_s regarding clamping on the threaded section. With a quill stem the force is effectively is transmitted to a point further down the steerer (where the expanding wedge is) and along the length of the interface between the stem and the steerer tube. When you use a threadless stem, that force will be located in the clamp area. The grooves of the thread on the steerer will effectively act as a stress raiser which increases the likelihood of a crack forming (or rather, reduces the force required for cracks to form due to materical or surface defects). Cycling will create cyclic loading at that point propagating said crack until ultimate failure.
andrew_s has pointed to an example of this failure. My comment comes from having trained as a mechanical engineer. I am definitely no expert in material fatigue and certainly a little rusty but I doubt anyone who was would be comfortable clamping a threaded steerer.
Out of interest, why are you reluctant to use a threaded to threadless adaptor as suggested by fossyant and myself? It seems like a relatively inexpensive solution that should prove safe and reasonably elegant.
EDIT: Just a quick note, the groove presumably runs perpendicular to the threads. This won't create the same stress concentration due to it's orientation and that it is not continuous around the circumference of the tube.
