But… in the example above all (or most) of the meshes are visible. So why is it then more effective to use static meshes?