What LCA boundary catches the most bad CCU claims?

With CCU, the same process can look useful or useless depending on the functional unit and reference system. That is not a footnote; it is often the whole result.

The two von der Assen papers are good starting points: doi:10.1039/c3ee41151f and doi:10.1039/c3cs60373c. Tanzer and Ramirez is the one I would use when a claim says "negative emissions": doi:10.1039/c8ee03338b.

If you could force every CCU claim to show one LCA boundary choice up front, which one would it be?