A citywide sample is probably too broad, but a very narrow neighbourhood sample may leave too few useful comparisons. Neither gives much confidence in a 28-day headline figure unless condition and occupancy are also controlled.
I’m also not convinced the slow listings should mainly be...