It's not really necessary to have even numerous parallel timelines. Perhaps 4 or 5. Say 2 or 3 where Romulus is destroyed, one where it is stopped, and maybe one where the events never arise.
But that is exactly my point. With so few timelines, the "randomness makes any nonsensical thing equally likely" argument is a complete non-starter, because that argument is predicated on the assumption of an infinite multiverse where everything's got to happen somewhere. With so few timelines, the probability that Romulus could be destroyed by two different supernovae at the same time by completely different and unrelated causes is just too ludicrously tiny to contemplate. This is elementary probability here. If you flip a coin a million times, maybe it'll land on its edge once, but you'd be crazy to expect it to happen if you only flip 4 or 5 times.