There's no such thing as the distance the light travels. There isn't even such a thing as the time the light travels. What there is, is two events-- the emission, and the absorption. How much distance, and how much time, that relate those events are purely given by the coordinates chosen, which presumably are in turn connected to the observables made by a given observer or set of observers. The sole universal constraint is the causal connection-- the absorption happens after (or essentially simultaneous) with the emission. The rest depends on the observers, but we can make certain assumptions to reduce the complexity of the "tower of Babel" of possible distances. We can assume the observers are "co-moving" with the prevailing mass flow of the universe, since most observers would be. Then we can sort those observers based on various distance measurements, like the cosmic time between emission and absorption (the "light travel-time distance"), or the angular diameter distance. The observers then collect on spheres that get the same answer to these various measurements. These spheres give meaning to the distances, but remember, it's all the observers. Absent of observers, distances have no meaning at all, and all causally connected pair of events can certainly be regarded as happening at the same place, separated by zero distance, just as any causally unconnected pair of events can be regarded as separated by zero time. And when those events are the emission and absorption of light, we have the singular opportunity to associate both zero time and zero distance between the events-- it is a "null geodesic". But we would generally not do so, we'd instead pick some observer and use coordinates that make sense to that observer.

It might seem like a nitpick, because we do talk about the distances to things, and how long ago they occured. But I still think it is important not to fall into the illusion that these are absolute statements of truth about the events themselves. It's all quite subordinate to the choice of observer-- generally, us. The answers should not have to state that every time, of course, but when statements like the above are made, I think it's time to set the record straight.