Thanks for heads up, we can try this material again later.
I was just looking at the ref photo, and saw these diagonal highlights which look like secondary caustics, which don’t occur in our renders. It’s probably because of the larger radius in the real hub compared to our model. Or it could be that we need UV’s to map the anisotropy correctly, I don’t know if the hubs are uv’d.
Overall, that pic has turned out to be difficult to reproduce, not just because we must guess on the material (and the geometry, which I drew in rhino), but also because I lack information on the actual environment, and don’t know what has been done to the photo post-wise.
I am really not even sure how much aniso the hubs have on their exterior — clearly they look to be as-machined on the inner surface, but I cannot tell how much polishing has been done on the outer, before application of the powdercoat. In addition, much of what appears to be due to anisotropy in the sharp vertical reflections, I believe is (partially or mostly) due to the sun angle, and the material of the table. However on the other hand, the highlight you point to above could not be coming directly from reflection off the table, except if the hub is not sitting level, or the table is not flat, so it could be from less-complete polishing on the flange.