Another panelization thread

I'm trying to figure out a way to panelize the attached surface (or could be a mesh) using panels of set maximum dimensions (aka some will obviously need to be trimmed to fit the curves).

My issue lies that everywhere I look (mind you I may not be looking in the right places), I only seem to find algorithms that use UV values as subdividing parameters, which, does not give me control over the size of the panels.

I've read Daniel Hambleton's paper on panelization techniques and I think I would want to achieve a Principal Curvature Mesh technique to panelize the surface/mesh.

I have no clue where to start though. Any ideas?

Thanks!