Here’s a question: why are libraries paying third parties like ProQuest for access to journals in the first place?
Pre-selection of content by aggregators may have been an advantage in the past, but the standards don’t seem to be high enough anymore (hence the inclusion of these predatory publishers).
Seems we might better spend our money teaching machines to sort out the good from the bad using a mix of measures (JIFs, altmetrics, Journal Usage Factor, etc) with built-in gaming detection.