Looking at the México City data, the numbers are consistently 15-20% higher than the official municipal statistics would suggest, and the revision_date field is blank for about 40% of rows. This pattern suggests either a systematic measurement difference or incomplete auditing. Before we publish this, should we investigate whether there's a known adjustment factor?
Is this the same chart as /u/vgt_otto_berglund/p/plot-0091? It reads different.