I just recently had a 992 (#2550837) serviced (cleaned, oiled and adjusted) by a local watchmaker who did a real good job, in my opinion. The watch was running and keeping good time within a minute or two of the U S Naval Observatory official time before I had it serviced, which I didn't think was all that bad for a mechanical watch from 1930.I would carry the watch during the day and at night I would place it horizontal on the nightstand with the dial up. After I got it from being serviced I checked it against the observatory official clock and it was about +20 seconds. Not a bad improvement. Then I thought to change the position of the watch at night to see if it would affect the accuracy any. So I went to placing the watch with the "3" up vertically on a little plastic watch stand and here are the results: after night 1 there was a change of -2 seconds ( +20 sec down to +18), after night 2 there was a change of -5 sec ( +18 sec down to +13), after 3 nights there was a change of -9 sec ( +13 sec down to +4 ). During each day I would wear the watch as usual. I have read from a book about Rolex watches that the position of a mechanical watch can affect the accuracy of it and here is the proof. I know that comparing a mech. watch-and from 1930 at that- to the atomic clock at the Observatory is perhaps asking too much but I am pretty pleased to have a 992 only +4 seconds off.

here is the reference from Rolex

I wonder how some of the watchmakers on the forum handle the situation where a watch doesn't keep particularly good time after servicing. There are many instances where a balance staff may have been replaced at some point, but not poised. Cleaning and oiling the watch may get it running, but it can take as long or longer to poise and time it.

Sorry if I hijacked your thread, Joe, but it seems relevant to your post.