I don't know if this asks the answer, but having myself either handled high-precision instruments before, and fiddled with inner rim diameter myself, before, I would consider these things:
- Even if you HAD the instrument yourself, in your hands, It would be actually difficult to measure, annotate and manufacture the dimensions precisely. I think even if you take the original to a metal-machining shop, some precision would always be lost;
- There would be no way to be sure anyone of us measure it right, even if we were most interested in this reverse engineering process (I am, for sure, but I don't have the tool unfortunately).
- Most times, reverse engineering is a process where reproducing the WORKING PRINCIPLE of a tool is much more important than reproducing the EXACT DIMENSIONS of the tool.
So, what I would suggest you to do is:
- Study and understand exactely what is the working principle of this tool. It seems to me it is to fit rods with nipple-shaped endings inside diammetrically opposed holes in a rim, and then measure the distance between the two other endings of the known-sized rods, so you can know some key distance, that I would define as the distance between the flat parts of the nipple's head. Having a spoke ending there would allow for maximum thread contact without the spoke thread prodruding towards the tube, also with a comfortable tolerance of 1 or 2mm shorter spokes being perfectly useable.
- Considering this working principle, take the most similar things possible: actual spokes with glued nipples in the thread ends, and some precisely placed measuring zones in the elbow side:
- The elbows themselves. You could tie them together, measure the distance, and add this distance to the already-known rod lengths;
- Some smart-shaped-object, like a piece of a metal ruler, rigidly fixed (glued?) in place of the elbow;
- The naked tips of the wires after removing the elbows.
If this is done thoroughly, I think it is perfectly possible to achieve, with no great expense and no great work, a level of precision more than anough to build a wheel within the relatively loose tolerance allowed by spoke/nipple thread lengths.
I know this is not the answer for your question, but I think these considerations might be of your interest.
Hope this helps!