I have a tangentially related question: has there been any thought given to raising patent validity standards around teachability? A big constitutional point of a patent is to teach someone "skilled in the art" how to replicated the subject matter of the patent. That's typically intepreted as "can you pay an expert witness to claim the patent teaches them to construct the claimed infringement," which is, of course, always true. But to me, this seems like fertile ground for the defense, and ought to be a negative test: can the defense find an expert witness to say that the patent is too vague, unspecific, or abstract to teach them to construct the claimed infringement.
Is this approach ever tried in these cases? It seems to me that the more convoluted and abstract the patent (which seems to be the point of many of these patents explicitly written to try and reach as broad as possible), the more juries will tend to identify with the expert saying "I can't make heads or tails of this thing" and invalidate bad patents as simply failing to adequately teach the construction of the infringement claim.