This matches my take. The problem is that these companies are serving two distinct roles: 1. Knowledge dissemination and preservation 2. Community consensus

This matches my take. The problem is that these companies are serving two distinct roles: 1. Knowledge dissemination and preservation 2. Community consensus and reputation building SciHub is the obvious answer to problem number 1. It costs literally nothing to distribute this knowledge to the community, and I'd call attempts to recreate the distribution channel foolish and misguided (at best). Number 2 is much harder. When I was in academia (and watching my wife go through a Masters and then a PhD in a science related field) SciHub was an excellent way to find and read a paper, but it was not an excellent way to judge the quality or trustworthiness of that paper.