if you commissioned the right people especially with modern AI scanning of originals.
Back when I was taking Jewelry classes in the 1990s I visited IWC. Had I been more successful they were interested in my efforts. They had a huge bone pile out back with the rejects. Most of this was ceramic dials. It is really difficult to get the basis even and uniform. I did have a PAL VHS tape of the process.
From my experience this is something AI would be quite terrible at. AI trends to more one size fits all generic abstract problems. Still have not bothered to try a chatbot. I have more targeted stuff with the player piano rolls. Most of what seems to be trending is smoke and mirrors. A lot of it is the dragon eating it's tail.
In the 1990s the time I was working with scanners and printers. So had access to some fairly high end tools and equipment. Silk screen equipment and toner transfer decaling paper. The same stuff used for putting foil logos on coffee mugs.
The short of it is dials are made with the same gravure process currency is printed with. It really comes down to the engraving on the plates. Which is an analog process. Modern equipment is a digital process where the tool paths are made from dots or short line controlled by servos.
Modern printers and scanners also contain anti counterfeiting measures. Some are more public than others. Pattern recognition has been around for a while. Another reason AI might fail as the 'security measures.' might train it to reject corporate IP.
A few weeks back I was thinking I could etch glass with the fiber laser, but in research the frequency is wrong although the laser does have the 360 degree servo axis. The CO2 laser might be better, although the owner does not recommend it and ask we not do it.
I have seen the plates and the steel is polished to a mirror finish. I have yet to do that as it gets boring really fast.
Spindles need to turn at high RPM. I broke the tips of most of my 60 degree engravers. I think I only have one left. I started to make a new spindle and never finished it.
Most refinishers are looking for a quick buck. There were still a few around in the 1990s and the quality was terrible. The ones who are decent are more like counterfeiters which is essentially what they are performing. There really is no value in spending time on dial painting due to the modern ideas of branding and IP.
Eventually I may get re-inspired to do more on this. At the moment I am trying to see IF I can simply cut some hands. I did burn through a bunch of brass making blanks. To be useful these need to be domed by spinning or pressing.
The YT vids of using the fiber to make plates said that some were taken down due to trademark infringement. So they re did them with a house branding and a 15 minute tirade at the start. They also show a pile of rejected blanks.
More evidence that the cost of replication in hours and materials will be more that finding an existent NOS dial of accepable quality.