You know, I've seen this discussion carried out any number of times, and the math just doesn't convince people this is impossible.
So, try this on for size: If this did work, it would be easy. It is every bit as easy as the idea sounds. Any lab set up for "quantum" experimentation could do this; entangling two things is step one in any experiment that you could call "quantum".
Yet this is not established, easy technology, with YouTube videos showing you how to set up your own FTL communication with your buddy on the other side of the planet, and off-the-shelf FTL networking equipment available for any ol' hedge fund or ISP who wants it.
This is because it's impossible.
If you're interested in why it's impossible, feel free to check out the many and abundant explanations of why it doesn't mathematically work.
But in the meantime, consider that if it hasn't been that commercialized, maybe that's because it's impossible. Because if it were just as easy as "entangle two particles and then poke one of them to send a message to the other", this would be trivial stuff. Nothing like quantum computing and its need for extreme isolation, this would just be a simple variant on stuff that really is off-the-shelf tech for quantum key distribution: https://infogalactic.com/info/Quantum_key_distribution#Quant... If FTL communication was just a matter of perturbing entangled particles, any of these existing, real-world, you-can-touch-them network setups could be turned into FTL networks with just a few small tweaks. No problem at all.
Obviously you're right on the physics, but this is like saying that supersonic flight must be impossible, otherwise we'd all be commuting that way.
If you can send a regular signal around the planet faster than you can read a quantum entanglement state, which seems entirely likely (to say nothing of bandwidth), then the technology would never have a useful terrestrial application.
This is why I hammered so hard on if it worked, this would be easy.
Colonizing other planets may or may not be possible, but if it is, everyone expects it will be very difficult. If supersonic flight is easy, it's obvious that it isn't necessarily easy. (It is, of course, possible, and it isn't easy. It's solved, but it's not easy.)
If FTL communication was just a matter of entangling particles and poking them to collapse this way or that, thus sending a message on the other side, it would be easy. It would be a lab demonstration in every college-level quantum mechanics class, right next to the double-slit experiment.
"If you can send a regular signal around the planet faster than you can read a quantum entanglement state, which seems entirely likely (to say nothing of bandwidth), then the technology would never have a useful terrestrial application."
If you haven't read those links ajuc linked in this conversation, you should. It's a very similar idea, getting to QM by taking superluminal relativity seriously. (I suggest this as an "interesting followup" to you point, not disagreement.)
Still, given that the "entangled state" is likely to be as easy as detecting polarization on a photon, since that's the whole point, it doesn't seem likely this would be the case. More likely the case is simply that FTL communication is impossible. You can send an entangled state around the world and read it in two places, no problem, today, plenty quickly. That's how the quantum key distribution works. You just can't communicate with it because you have no influence over how the entangled state collapses.
(In fact, there's a sense in which quantum key distribution works precisely because the properties you'd need for FTL communication don't exist. If they did, quantum key distribution wouldn't work safely!)
Entanglement "simply" means that the other party will have a correlation with your measurements.
Without receiving information of what you measured (inevitably in a slower-than-light way), none of their observations can tell whether you did that thing or not. The behavior of a single entangled particle is perfectly indistinguishable from a non-entangled particle, it's just that for entangled particles certain combined results are impossible or unlikely.
Imagine being given a magic pair of dice which always adds up to 7 when thrown at the same time. If you take them to different rooms and get a 6, you know the other party should have gotten a 1, but you can't use it to transfer any information to them because no matter what you do, in isolation their results are indistinguishable from normal dice.
To my understanding, quantum is nature's method to some things we have had to derive artificially. Entanglement does not involve "communication". Changing the spin on one particle does not change the other in reverse. When entangled, ones spin will oppose the other, but you can only ever know what spin that is by measuring it (with scientific instruments), which "breaks" entanglement, therefore you can't do anything further to it. This is one of those problems where you're working with the very fabric of the universe, and there's no debugger. You can't know the state of the particle without measurement, but measurement changes it.
Both of those options have the same measurement statistics on the other side. You can't control whether the other side sees On or Off via entanglement.