I'm not sure it's really being thought through so much as an emergent effect of the drive towards metrics. Universities want to (or are pressured to) quantify their researchers' productivity, which results in things such as impact factors, h-indices, number of papers in top-tier conferences/journals, and so on. Researchers then have to maximize these metrics.