It's the same "discovery problem" that every long tail market encounters: a significant portion of the tail is going to be either low-quality or irrelevant to m

It's the same "discovery problem" that every long tail market encounters: a significant portion of the tail is going to be either low-quality or irrelevant to most consumers of that domain. So you need aggregators, and those aggregators will face pressure to moderate the long tail by surfacing the most relevant results, while also still providing access to the tail in some form (see also Ben Thompson's aggregation theory [1]). Unfortunately, this dynamic is more social than anything: there can only be so many aggregators that can occupy the zeitgeist. That scarcity of attention leads to rent-seeking behavior, which is the main problem with academic journals. The current situation is probably approaching optimal; large institutions fund the moderation and aggregation through journal subscriptions, while access is widely available via Sci-Hub. Without a profit motive, such moderation would be even more politically-driven than it is today. At the same time, the Sci-Hub community does need to ensure its continued existence because it is such a massive boon to science writ large. [1] https://stratechery.com/2015/aggregation-theory/ https://stratechery.com/2015/aggregation-theory/