Emmanuel,
Thanks for your reply. When I said we may only 'need' GR for highly curved situations I did not mean that it wasn't useful for things like GPS, only that GPS can probably be handled with 'weak field' approximations. And that was my point. As Sweetser shows, we can describe the world via potential in 'flat' coordinates or by distorting the coordinate space. In 'weak field' cases (such as GPS) the choice may be optional. Or not?
We agree that it's good to be speculative in approaching physics. Because QCD generally yields 4 or 5% accuracy, QED lately yields 4% accuracy on muonic hydrogen and is off by 120 orders of magnitude on vacuum energy, and GR is off by 0.2% on LAGEOS pericenter precession, I believe that conceptual issues, not just mathematical issues, need rethinking.
I also offer a number of very specific examples in my essay that serve as 'test cases' for theories, and that, so far, no current theories can explain. This is, in my mind, more appropriate than speculating on theories based on Planck energies that we will probably never reach. I believe that it is physics when we treat real known anomalies that demand explanation, and mathematics when we treat speculative scenarios which may never be achieved. Today, things seem more focused on the latter.
Finally, you state that Bell tests show hidden variables to be false. Based on Anton Zeilinger's excellent book, Dance of the Photons, I believe that this conclusion is premature, as I explain in my essay.
I agree with the logic of your last statement concerning QM and SR, but do not yet see how photons alone can satisfactorily evolve our universe.
Thanks for your perspectives.
Edwin Eugene Klingman