I'm not sure if this set-up has been put forth already but I'm curious what people think of it. It is about ascertaining the correlation of the polarization state of a pair of photons(A and B) by two observers considered inertial (x and y)that await at the same sufficient distance at opposite sides from an experimenter that sends two entangled photons in opposite directions. Each observer upon measuring the polarization state of each of the respective photons A and B automatically sends the result to the other observer electromagnetically. So for each observer x and y we consider two events timelike separated: measurement of one of the photons and reception of the info of the measurement of the other respective photon. Observers x and y can be considered spacelike separated regarding the physical process of correlation of this particular pair of photons they must verify and with respect to wich they act as observers. Each observer verifies this physical process with a pair of timelike events that is temporally reversed with respect to photons A and B, right? Also these events observed by each observer are inside the future lightcone of the event of the experimenter sending the photons.