While I can and do access materials from libraries, including online access, Sci-Hub is both more complete and far more reliable and convenient. Find a resource, plug in the URL or DOI, and I've got it. Versus locating the same reference independently through one of several distinct libraries, each with their own multiple subsystems, authenticating, and sometimes, sometimes not, securing the material.
Another point Bohannon failed to address, which is covered in the discussion here, is the role of journal publishers as gatekeepers not only to content but to careers. Academics, increasingly squeezed by budget retrenchments and awful working conditions[1], must publish through prestige journals in order to establish and advance their careers.
Journal publishers are rent-seeking at both ends of this channel.
Sci-Hub, or as I like to call it, the Library of Alexandra, hs a tour de force demonstration that information is a public good, and that information access wants, and needs to be free. Sci-Hub isn't a complete answer to the problems of current academic publishing (again: publish or perish), but it's a relief valve for many, and an absolute and irrefutable proof of the pressing demands for access.
________________________________
Notes:
1. See the amazingly awful story of a young newlywed biology postdoc who lost her arm in a lab explosion involving an improperly instrumented gas cylinder in which oxygen and hydrogen were being mixed under pressure. This after repeatedly reporting short circuits and electric shocks from the equipment.
http://chemjobber.blogspot.com/2016/03/postdoc-loses-arm-in-...