Smart idea, that mirror.
Thinking back at the twist angles in the dye penetration experiment, I now realise that the difference in twist angles between your measuring in the spinning and the resulting twist in the fabrics is probably due to the weaving process. Because we had to cut and rethread after every 25 cm we lost twist. If I could have woven on until I reached the lenghts of yarn packed on the warp beam, those probably would still have had the original twist angle. I`m making Christmas cards with the overtwisted singles green cotton and untangling the warp bundles for threading cost me quite a lot of the extra twist I put in.
The twist in the finished fabrics is actually not so much lower than what I aimed for in the spinning. It is quite irregular, though, which might be due to the spinning technique I used or to the high speed I tried to achieve, or both. I had a helpful colleague who gave me her readings of twist angle on a few of the hard-spun samples, and she read out angles of about 40-50°, so right where I had been aiming at. Which is nice