The Novikov self-consistency hypothesis is pretty neat. I like the thought of the researcher travelling back in time to create the apparatus such that the result always confirms quantum entanglement. It's classical and not quantum so it's not clear to me how it relates. I'm not sure anyone (besides philosophers) takes the loopholes seriously.
Under this hypothesis, even if entanglement does allow changes to move faster than light--essentially travelling backward in time--any action that would expose this time travel would be forbidden. The scientist is physically prevented from making the "wrong" choice when setting the detector.
The problem with this hypothesis is that it seems essentially untestable.
Another way to think of option 2 is that the universe is 100% deterministic. Under this hypothesis, the speed of entanglement is only apparent, not real--like planning with a friend to both turn on your flashlights on opposite hilltops 2 hours from now. Without knowing that it was planned in advance, an observer would wonder how you reacted to one another so fast.
This is also impossible to fully test; at best you could prove that the consistency extends as far back as we can see into the past. And that is exactly what these scientists will be testing.