Performance issues with vtkOpenGLBatchedPolyDataMapper

Yes, that makes sense. the mapper builds different buffers based on those settings. Have you considered upstreaming this?