Hi Ben,
Thank you so much for your encouraging and insightful comments. I'm so glad you share my interest in the profound significance of covariance and observer-dependence/independence. The history of physics seems to suggest that separating the invariant from the observer-dependent is the key to getting at the true reality beneath, and I'm fascinated by the ways in which quantum gravity undermines invariances that even relativity and quantum mechanics had left intact.
With regard to your comments:
1. I agree that the question of how to define global or even local observables in quantum gravity is extremely important and mysterious. Personally I am intrigued by the notion that while for AdS or asymptotically flat spacetimes you can retain some kind of invariant boundary observables, you can't seem to do so in de Sitter space, precisely because the de Sitter boundary is observer-dependent. This to me is suggestive that reality is far more observer-dependent than it seems.
I've just read and greatly enjoyed your eloquent essay - though I think I'll have to read it a few more times to understand it! (That's a reflection of my nonexistent mathematical background, not of your essay!) I'm curious if it is in any way related to Tom Banks's work on holographic spacetime? As I understand it, he argues that the causal structure of spacetime can be reconstructed from quantum commutation relations up to a rescaling of lengths and times, and then you can use the holographic principle (because it gives you an area as a function of the number of quantum states) to include scale and now you've got spacetime structure. I believe in his work the observables are noncommutative matrices on the boundary of each "reference frame". Sorry if that's totally irrelevant. In any case, I'm in full agreement that the manifold won't survive quantum gravity - the dualities of string/M-theory certainly point in the direction of a kind of emergent spacetime. In your model, with its basis in binary elements, would you say that the world is "made of information"? And is that information observer-dependent?
2. You're absolutely right - I probably should have mentioned that there is disagreement about how much we can extrapolate the lessons of black hole horizons to cosmic horizons; even Susskind himself has gone back and forth over whether horizon complementarity applies to the de Sitter horizon. Personally, however, I'm unconvinced by arguments that they shouldn't be treated equally. They are mathematically equivalent, they share the same properties of entropy, temperature, etc... yes, there are physical differences (the dS space doesn't "evaporate away", etc) but it just seems to be telling us something general and profound and as a structural realist I see every reason to treat them as equivalent - not least of all because of the equivalence principle!
Thanks again for your comments and for your fascinating contribution, which I will now go and re-read :)
All best,
Amanda