Thankful for being wrong. My previous mobility graphs had an axis error (time on the mobility axis was misaligned - shame on me for doing these in a spreadsheet): mobility has been slowly bending down for the last 5 weeks. Is that good? Yes. Am I super excited? No. Here is why:

I was quite worried cases kept growing while mobility did not seem to have budged. What was good from that perspective was that it suggested there was a lot of room for improvement. The problem is mobility has been slowly decreasing, while cases keep going up. Is that bad?

Hard to tell. While it's really good that the population changes behaviour in response to rising case counts (before the government acts), cases continued rising. But, maybe not as fast as before; see slight bend in the trajectory of cases ~2 weeks after decline started.

There are many factors to take into account when interpreting ecological level data + impact of the changes in testing criteria affecting the October case counts. It is possible that what we see below (change in slope of 7-day mean) suggests that lead to missing ~200 cases/day.