[palmtree3000]

"Incorrect" is way too strong here.

I agree that a reference frame that has about the same velocity as the stuff you care about is in some sense privileged. But the fact that we haven't seen the merger happen yet means that it's outside our light cone, which means there exists a reference frame where the merger hasn't happened yet.

that is to say, t_{this_coversation} < t_{merger}, in that frame.

[andrewflnr]

It is not too strong. GP specifically claimed that simultaneity depends on "where you're standing". This is wrong in all reference frames.

As for timing of distant events, GP didn't make any reference to other reference frames at arbitrary velocity, only different positions. In our reference frame, which they were implicitly talking about, the merger (assuming the prediction is correct) was in the past, unambiguously.