I seem to remember reading about this some time back. Hopefully Polymath will chime in, as he has far more expertise than most, on the subject.
I also seem to remember reading about quantum events, where the "cause" happens after the "event" in question, a kind of causal reversal.
As the famous physicist, (Feinman?) quipped: "If you think you understand quantum mechanics, you don't."
All the problems people have with quantum mechanics ultimately come from wanting a classical description involving particles with definite properties and at definite locations at all times.
But quantum mechanics is an inherently probabilistic description of things. It is not, ultimately, a causal description. yes, things we do will affect the *probabilities*, but we cannot determine which specific property will show. We can only determine the probabilities and correlations between those probabilities.
In the 'causal reversal' experiments, there is no actual reverse causality. There are correlations and those correlations proceed forward in time at all points. Furthermore, the correlations do not transport faster than light.
So *any* claims of faster than light communication from entangled particles is simple garbage and a fundamental misunderstanding of what is going on.
Similarly, in the scenario of the OP, the correlations are moving forward in time and the interactions between 2 and 3 affects the correlations. So, because 1 and 2 were correlated, and since 3 and 4 are correlated, any induced correlation between 2 and 3 produces a correlation between 1 and 4.
Like I said, you get problems when you try to describe quantum mechanical systems in terms of classical notions of particles. But, quantum mechanics is the description that *replaced* the classical one. So attempting to have a classical description is to attempt to describe the system with the old, outdated and *wrong* descriptive apparatus. So, OF COURSE, you will have troubles.
Do I think there are interesting metaphysical issues here? Absolutely. And one huge lesson is that classical metaphysics is simply wrong.