[taliesinb]

Information transfer was never a problem with spooky action. But if all entangled particles were linked by some kind of funky wormhole it seems a harder task to explain why FTL information transfer isn't routinely possible.

[jerf]

All of the arguments about why it's not possible remain true, even if connected by some sort of wormhole that means they "really are" right next to each other in some sense. You still have the fundamental problem that you can't select what you collapse, so you still have to send a conventional signal to "decode" the putative instantaneous signal.

Remember the correspondence principle will still be in play; future quantum theories will still have to limit out to what we know today, because the QM of today is arguably the most rigorously tested theory in humanity's history, by number of significant digits. FTL communication still will have all of the problems conventional relativity says it will, for all the same reasons.

Entanglement has suggested for a long time that there may (in clumsy English words) be some sort of "real reality" that isn't necessarily constrained by what we think of as space and time. In fact even relativity looked at in a certain manner has suggested this; you can travel from any point in the universe to any other, barring black holes, along null spacetime intervals. Null spacetime intervals have no distinction between the points in them, because they all come out the same 0 in the metric measurement. I wouldn't expect that any of this new math is going to change anything; it may explain where the constraints of space and time come from, but explaining the constraints doesn't mean that the explanation will come with a way to get around them!

In fact my personal observation is that the fundamental limits of space and time have been getting stronger as we learn more about physics, not weaker; a mathematically rigorous derivation of the fundamental speed-of-light limit from a "more fundamental" level of reality locks the door even tighter, it doesn't open it.

[tfgg]

Indeed. When I saw a novel proposal for embedding QM inside GR using closed time-like curves (rather than the usual vice-versa), the main problem seemed to be that a GR-computer is potentially _much_ more powerful than a QM-computer, so you have to explain where the extra power has gone.