Yes, the problem solved in the paper (Tower of Hanoi) is far more easily defined than 99% of actual problems you would find in commercial software development. Still proof of "theoretically possible" and seems like an interesting area of research.