This article would do well to have a line graph included
You mean like this? So many people were repeating baseless claims I started my own experiment last night. The graph is power in % versus time in minutes. Each point is an individual time collection. The lines are linear least square fits. It was not a rigorously controlled experiment as I was basically going about life. But it does offer more time points.
For 15 hours I had it sitting out of the case on a table. It was connected to anywhere from 3 to 5 devices, but I was not using it. In that time it consumed 5% power... when you least squares fit it, the loss is about 1% per 3 hours. (the blue line)
Then I used it for 2 hours listening to a variety of things, you can see the steep decline marking faster power consumption, or about 5% per hour, or 1 % ever 11 minutes. (the orange line)
Then I took it off and let it sit again. There was a stepwise function that was strange, it seemed to go immediately into a standby mode, very little power drain for 15 minutes, and then it lost in a very short time, a few minutes, 2%. I am not sure this is real, it may represent the device battery recalibrating. In any event, after that observed drop, it started using power more slowly again, roughly at the previous slow rate, though I am going to collect more time points to confirm. (the grey line)
I have not tested it in the case for a long time, I can confirm it disconnected from all my devices when I did.
Measurements of these types ideally take days, not hours, to get better accuracy. But I think the trend is clear... out of the case the device will drain in 10 day. Using it, the 20 hours seems reasonable. More work needs to be done to figure out how fast it goes into a lower power mode when you take them off, but it certainly is not as fast as using them as some you tubers said without researching it and people here parroted.
Thanks MR for actually doing a test and giving us this data. I look forward to seeing (and sharing) more.