I do wonder what the idea was of averaging by month when reviewing by year of readings. That happened with weight (with the absolutely regrettable change being the average per year when looking at all the tracked data over the different years). And now also with health statistics. I found the trend per 30 or 90 days many times more useful than the average per month when looking at data per year. How do these changes work? Is there a user panel or evaluation moment somewhere among the users to see if they see it as an improvement as well?
