It happened sometime between whenever we observed it and 2600 years ago. It's equally valid to say that light travels from there to here instantly (but takes twice as long on the outbound trip) or the reverse. Taking the average is just a convention.

Hence why the articles title - which is based on when the light cone of the event reaches us - is actually the better way to think about it. At least there's no "depends how you feel like defining the speed of light today" in it.