So I'm working on my FBX importer/wrapper. I understand how shape-deformations are carried out (just multiple meshes that are blended to obtain new shapes, etc) I'd have to move these calculations to the GPU, I guess create VBO's for each of the shapes, and do the calculations in the shader, etc. But is it worth the effort?
The game I'm working on is a fighter. I expect lots of skeletal animations and some animation blending. Don't really know when blend-shapes are useful or appropriate!