Last night I found a sign error which occurs twice in Section IIC of both the current arxiv version (v14) dated 25 June 2019 and the slightly revised version of 21 October 2019 posted on Academia and Researchgate. The error invalidates that subsection but not the entire paper, although it seriously hurts the persuasiveness of it. It will take me several days to a few weeks to get a revision posted. I will try to save that section for a little while before I can give up and delete it, which may force some narrative rewrites as well. So far as I can think of it shouldn't propagate as math errors into the later sections, which were developed prior to that section.
To be clear, it is this 25 June 2019 arxiv version: https://arxiv.org/abs/1609.04446v14, and this 21 October 2019 version posted here and here. These all have the title, "Spin-Spin Coupling as the Origin of Atomic Binding."
In the arxiv version, the error is in the unnumbered equation priorr to Eq. (29). This is a 4-step equation. The sign changes at the third equality. It should not change. That step accounts for a vector triple product which adds a negative sign, but also a substitution for a_t, which also adds a negative sign. The latter was apparently overlooked. In the Oct 21, 2019 version, the same error occurs in the unnumbered equations prior to Eq. (33) and Eq. (34).
Chex n balances
4 hours ago