And that's not taking into account that much of academia comprises disciplines where nothing is exactly "useful", or expected to be (e.g. arts and humanities). Metrics of worthiness in these areas are naturally going to be pretty subjective. (Which is not to say they couldn't be crowdsourced.)