Fini, I think you actually put your finger on a problem with both. You have pointed out a tautology present in both AA and AVRT . With AVRT “judging if the decision/commitment was "real" by the result” and again with AA by judging if the steps were “worked” sufficiently by that same result. The result defines the adequacy of process that previously took place. With such a definition there can be no “failure” of either method.
The trouble with turning a genuine failure into a retroactive definition of an inadequately followed program or method is that we learn little or nothing in the process.