When you do this I'd love to read about it.

You're right if you're aiming to make a 16th century pocket watch. Anything approaching a modern wristwatch calibre is much more time and money intensive.

The majority of the cost will be time and effort. I think George Daniels estimates it to be 3000 hours for a skilled amateur to make the tourbillion pocket watch described in his book. This number will be decreased if you forego the tourbillion. But even so, other equipment costs: jewelling press, staking set, poising tools, depthing tool, all kinds of abrasives, oils and greases. The largest tool cost will be lathe attachments like collets, cross slide, milling attachment, wheel and pinion cutters. You may be able to cut wheels and pinions on the pantograph but probably not to the precision needed for a wristwatch, maybe for a large pocket watch or clock. Maybe you could put the lathe on the milling table to index the stock and use the mill to cut?

That's not counting all the theory needed to design and build a movement from scratch.

My advice would be to steal the going train, escapement and balance from an existing movement and fabricate the remaining parts

A good starting point is to try to build a pin-pallet escapement design, like Timex used in the 60s and 70s. This is largely doable at home with a 3D printer and basic tools, e.g.: https://incoherency.co.uk/blog/stories/the-watch-project.htm... Getting to modern accuracy is hard though.