Hmm, sorry I can't be explaining myself correctly.
Let me try again
Currently Blender only allows sub surf options on a per mesh basis.
Now lets say I am modelling a head as simple example. I am working on the eyes, eyelids, and eyeballs and am using sub surface modelling. Now to get the eyeballs to fit in the sockets correctly, I will need to turn up the subsurf value for the interactive display for the eyeball and the eyesocket (head) to the same values I intend to render them in. And of course the eyeball and the eye socket dont need to be the same value.
In this example the head is attached to the rest of the body. (same mesh) so posing with armatures becomes sluggish due to the subsurfacing.
So I select the body and turn down the interactive subsurfacing. For optimum speed I select the eyeball and turn down the interactive subsurface value there too. Etc.
Now I pose the model. Lift an arm and now I would like to see how the mesh is deforming so I select the body mesh again and turn up the interactive subsurfacing. Looks good, hey but now from this pose it looks like the eyesockets are not quite right. Select the eyeball and turn that up.
With more detail. (Hair, accessories etc) this becomes quite tedious.
One button to turn down all the meshes interactive subsurf level in one go and then turn them all back to their original settings again when necessary.
Its not quite the same as having them all turned down for interactive and all set properly for render. rendering is one way to see yes, but doesnt help when modelling.
Is this clearer?
------