I can't how that'd be an improvement on just replacing the figure. Which we can already do right in Poser. I mean, if it's going to need a different .obj file then the existing way is already faster than anything I could script up in Python.

There may be SOME simple optimisations that I haven't done yet in tinkering around with old half-forgotten code, but beyond that it'd have to be newer cleverer optimisations that will make my brain hurt to attempt - or be in native code as part of Poser (in which case that's when I'd be more tempted to look at expanding the capability to being able to have different uv seams/uv counts).

Cheers,

Cliff