I have four watches with the ETA 2824 in them now, so I decided to do a power reserve test on the three of them to see how much variation there was within the movement. I wound each of them 40 turns of the crown from a dead stop, set them to midnight, and let them run down. The variation surprised me:
Steinhart Ocean Black DLC (-5 sec/day) - 37 hours
Halios Laguna (-10 sec/day) - 41.5 hours
Halios Puck (-10-15 sec/day) - 25 hours
The parentheticals are the typical accuracy of each watch when being worn for a day. None of the watches had been worn in a few days.
Is this much variation surprising to anyone? The Laguna running that long was interesting, but the fact that the Puck only made it a little more than a day was shocking. I'm going to re-run the test again tonight so I can have averages. It would be neat if people with these and other 2824s could do the same test, so we could gather the data to establish a distribution of power reserves for this movement.