In watchmaking, the word «finishing» doesn't describe a single technique: it's an umbrella under which operations coexist that have almost nothing in common except their goal — giving a component the shape, finish or decoration specified by the design. From the bench I've learned that the first mistake is using it as a synonym for «quality»: a well-executed CNC finishing is worth as much as hand engraving if it serves the purpose; the problem arises when the word is used to mask the absence of details.
Finishing operations divide roughly into formative, decorative and polishing, and they often overlap. On the formative side, milling a dial from a solid block of aluminum — as Leica does on ZM calibers — guarantees superior dimensional precision compared to assembling stamped parts. The same principle applies to sapphire: extracting a curved 1.2 mm case from a monolithic crystal exponentially multiplies complexity compared to flat geometry, because the material doesn't forgive even minimal feed-rate errors. Tantalum, nearly double the density of steel, requires completely different tools and cutting speeds than stainless steel, and its dark gunmetal finish allows no corrections: every tool mark is permanent.
On the decorative front, the distinction that really matters is between repeatable mechanical finishing and irreproducible manual finishing. Engine turning and guillochage en ligne are executed on pantograph lathes or rose engines: they produce precise, replicable geometric patterns, but still require an experienced operator to calibrate the machine and manage pressure. Grand Feu enamel is a special case: every firing above 800 °C is a risk, and the percentage of rejected pieces on hard stone dials — lapis lazuli, jasper — can exceed two-thirds of the total. I've seen lapis lazuli dials crack during secondary hole drilling, after hours of work: the stone gives no warning.
Skeletonization, which Audemars Piguet calls openworking, is technically a formative operation on the movement: material is removed from the plate until only structural sections remain. The visual result is secondary to the mechanical challenge, because every removed piece of material modifies the frame's rigidity and must be compensated during the design phase.
Finally, composite materials — IWC's Ceratanium, Super-LumiNova excluded, TiDamas, Citizen's Super Titanium — often emerge precisely to resolve a finishing conflict: Grade 2 titanium is more corrosion-resistant than Grade 5 but won't polish and deforms under stress, so thermal or surface treatments are developed that modify its response to the tool. Visible machining marks intentionally left on some cases aren't laziness: they're a design choice that openly declares the process.