[danbruc]

I have seen that your profiles says theoretical astrophysicist so you certainly know a lot more than me about the topic. Nonetheless I can not see why we can not have a reference frame extending over 100,000 light years or at least over 1,800 light years. That seem relatively small distance, the relative velocities are relatively small, space is relatively flat, expansion of space is, I guess, not really relevant within galaxies.

So naively I would expect that you can attach a reference frame to almost all objects in the Milky Way and they would agree on simultaneity to within hours or maybe days. There are certainly some notable exceptions like the central black hole or particles traveling at an appreciable fraction of the speed of light and you can certainly just invent a reference frame with huge relative velocity changing simultaneity a lot.

But are there really places within the Milky Way so that simultaneity would be off by hundreds and thousands of years over a distance of just 1,800 light years? If yes, what is the cause of that, as far as I can tell it would have to be an effect of general relativity because, again as far as I can tell, there is no problem of defining simultaneity across extended distances in special relativity and the involved velocities are not large enough to have an appreciable effect to begin with.

EDIT: Just calculated an example Lorentz transformation, 500 km/s relative velocity and 1,800 light years distance, and the time difference comes out at almost exactly 3 years. That is certainly more than I expected and adds up to 166 years across the entire Milky Way, on the other hand 500 km/s is probably quite a bit above common relative velocities within the Milky Way.

So I am still not convinced that picking a reference frame for the entire Milky Way, say with the origin at the center of mass, one axis coinciding with the axis of rotation and rotating with the Milky Way so that the angular momentum vanishes, could not provide a good enough reference frame for the entire Milky Way for back of the napkin calculations.

[dsqrt]

I agree that you can make such a construction in theory. However, in practice there is no way we can synchronize our clocks with an observer at the location of that binary star system. There is also no reason to do it: the precise time of the star merger does not have any physical meaning, it is just a label. Saying that some astronomical event took place on the date we observe it, is equally valid and free from ambiguities.