Ehh… I don’t want to get too much of a reputation as the splitting hairs over semantics guy, but sometimes splitting hairs is important, so here goes 😛
Sure, it is true that the data in hand does not say 28-30. Some assumptions about what the run looks like from now ‘til then have to be made to arrive there as a final range.
But I would say, it is equally not the case that the data in hand says 24-25. Some assumptions about what the run looks like from now ‘til then have to be made to arrive there as a final range. Maybe the latter set of assumptions is better, but an actual affirmative case needs to be made for that — using the straight t-x comp implicitly embeds the rest of run sales ratio being equal to the cumulative sales ratio at t-x, which is a baseline heuristic that is probably pretty easy to outperform with some other pretty simply heuristics that incorporate e.g. recent pace, length of sales window, and raw level of sales.
I understand the appeal of the t-x comp from a computational convenience standpoint, and I’m not saying they should be retired or anything, but I think it’s a bit pernicious as far as implying a certain number as a sort of “what it’s looking like right now” which is not necessarily what it’s actually looking at right then in any meaningful sense.