Some months ago, after some research, and after posting a thread in the winders & tools forum, and after reading others' thoughts, I purchased a very nice 4-watch winder from Brookstone.
The price was affordable; under $200 including shipping and an extended warranty. The product was quite attractive with a very nice walnut finish. I have no complaints with the device itself, but the following realization has led me to change my opinion on winders and their utility.
To wit: Of the 3 automatics that I have been wearing, each of them tend to lose several seconds a day when in the winder. Those same watches generally run from +.5 to +2 seconds per day when manually wound and left dial up at night. If I leave my Datejust, mounted on the winder, with the winder turned off, it also loses time. My assumption is that this is an issue of positional variation, and it is relatively constant between watches.
Off winder, the Rolex Datejust may lose about a second daily when worn, and gain a similar amount face up. Pretty similar with very slightly more variance with the Seamaster GMT (1128). The Glashütte Hand/date is too new to fully assess, but it too appears to be very accurate when worn or off the winder, and again loses time on the winder.
As a result of this, I find that I tend to keep the watches I don't wear regularly mounted on the winder, but with the device turned off. The watch winds down, I wind and set it when I wear it, and the winder becomes a fairly expensive display case. The watches I wear regularly are often kept face up in front of the winder at night time.
Any thoughts on this regarding positional variation, or a different factor that didn't occur to me? Should I increase the setting for TPD above the recommended settings? I wonder if a winder that kept the dial up would suffer less from this phenomenon, or would it just introduce different errors in different watches?
I realize this could have been posted in winders & tools forum, but I think the question may have some general interest, and is not about a specific product as such, but about the utility of a class of products.
I'm interested in your thoughts. Thanks.