After the exchange with @YinMills reproduced in the previous post, I felt a bit deflated to realize that GU is a kind of "gravi-GUT" (a realization which came when I was explaining my take on the theory). A moderately unusual one to be sure, with a number of unique features; and one might think that being able to relate it to an existing research program, would help make the theory more accessible.
However, I am also a gravi-GUT skeptic, as well as a skeptic regarding gauge theories of gravity in general. 3d gravity from Chern-Simons (as studied by Witten) seems to be the only one that works, and even there Witten expressed doubts that it's the full story. As for loop quantum gravity, for a while my take has been that the LQG mainstream doesn't work, but that the Ashtekar variables probably can work, so long as they are quantized in a way that converges on standard perturbative quantum gravity, a la John Donoghue. (Cheung et al's rediscovery of the LQG area operator is probably also a clue.)
For me, GU is interesting enough that I would want to develop my skepticism more rigorously. Meanwhile, it has those unusual features, among which Eric's use of torsion stands out. So another way to investigate GU, would be to look rigorously at the prospects of gravi-GUTs with enriched torsion. I have maintained for several years that one reason why GU is interesting is that it's not even clear how it fits into the taxonomy of known theories; thoughts like these are part of that process.
