Here are my results of calibrating the EPD over the following day ranges of run 22: 9-15, 44-50, and 58-64. These days were uncalibrated during the live calibration period so they have only now been completed. Some notes

For all days, the following tiles are dead:
1 2 1
1 2 4

For all days day 44 and beyond the following tile is also dead:
1 6 27

Results were split depending on tile behavior. Particularly, tile ADC shift occurred in several places during the 44-50 period, necessitating an almost bi-daily segmentation of calibration values. Other time ranges were more well-behaved; however, day 14 had no data and day 62 had no data, necessitating the splitting of those day ranges into two parts.

Calibration results and timestamps are available in the zip file.