Made fulfillment a M:M between economic event and commitment

History: We did a process model then an exchange model, and they fulfilled commitments differently, with exchange being a M:M, and process being a 1:M. This resolves that in favor of a M:M for all events and commitments, in the interest of simplifying the model. We had a lot of discussion about this, and I believe @elf-pavlik was in favor of this solution, while I was concerned about people fulfilling commitments across processes with one event.

I now would like to resolve and simplify, and think this is the safest and most flexible. (This is triggered by the VF graphql api I'm working on for NRP/OCP.)

I'll leave the PR out for several days in case anyone wants to comment.