Dear Maulik,
You are one of the few authors whose writing and science strike me as equally excellent. I thoroughly enjoyed reading your essay. A few remarks:
1. Generally, I tend to prefer "local" constraints to "global" ones (page 3). (The reason for the quotation marks is because ultimately I mean these terms in the interaction sense rather than the metric sense; obviously things like entanglement involve some form of metric nonlocality; in classical GR, this distinction is irrelevant.) Anyway, the reason for this is that I feel we should still consider ourselves fairly ignorant of the global structure of the universe. Whenever "local" conditions can rule out global pathologies, so much the better!
2. Some of the worst global pathologies (such as the grandfather paradox) arise from clashes between causal and metric structures. A number of ideas for theories of quantum gravity define away this problem by taking causal structures to be fundamental (causal dynamical triangulations, causal sets). I am working on another such approach myself, but I have a long way to go!
3. I suspect that it may ultimately be necessary to reinterpret covariance (pages 3-4) in terms of order theory (rather than local group symmetry). I am not yet sure what if anything this would imply for energy conditions; as you point out, the "core principles of QFT" don't suffice for this!
4. It's an important point that you can always rely on the second law, since its justification has nothing to do with any particular physical model (pages 4-5). I personally suspect and hope (optimistically!) that the second law will be at the heart of any successful theory of quantum gravity and fundamental spacetime structure, and may ultimately explain things like the dimensionality of spacetime.
5. I agree that the division between matter and geometry is artificial (page 5).
6. Another nice motivation for string theory (pages 5-6). My own approach is in a different direction, but this is good to know.
Thanks for the great read! Take care,
Ben Dribus