Slightly rework logic regarding when to switch to grouped mode of Text Only face

Now we do not try to show groups if no total sensors have been defined.