OK. Try it this way. We, almost, accept that angle of a spoke sets its tension. The smaller the angle the greater the tension. On my front wheel the sums, which may not be correct, show a calculated 17Kgf difference inner/outer on the flange. This is with a 3mm offset to which you have to add the diameter of the spoke as it exits the flange and account for the offset at the rim. As to whether it is significant or not it is there and despite questions about the accuracy of the measurement the relative difference can be measured and set.
Now we get back to this torsion, radial as opposed to lateral, thing. The torsion, should it be there, will act to rotate the hub in the same direction assuming the lacing forces this which, if you follow Sheldon Brown, it must do so. In doing so the radial angle, again angle is the important part, will adjust to equalize the radial tensions, as vectors X/Y, in the spokes whilst the lateral tensions, as vectors Z with the axle being the Z axis, will maintain their difference. The complete wheel will adopt a minimum energy solution and there will be no torsion in the hub.
The only remaining question would be are you bothered and if so is it valid to calculate the difference purely based on the Z angle. I am bothered because I can measure it and also I am bothered because it makes sense, at least to me, and removes some of the mystery or uncertainty. I also dropped quite a bit of money on the TM-1.
Last edited by chorlton; 11-19-22 at 11:42 AM.