That's the problem with metrics though. You can make anything sound plausible. Once a model has been presented to you, you can come up with an endless stream of

That's the problem with metrics though. You can make anything sound plausible. Once a model has been presented to you, you can come up with an endless stream of just-so stories for why it measures what you were told it measures. Nonetheless, some metrics simply aren't that useful, and get adopted anyway because they produce a bunch of data. Psychometrics like this have an awful track record for standing up to more meaningful tests like "Can we drive any outcome we care about with this"