Another article I read recently stated that it's simply much easier to use SciHub to search a vast trove of papers than it is using the individual tools/websites of those holding these research papers hostage behind their subscriptions.
Universities search (While crap) will be far superior than anything Scihub could ever ever pull off.
And as mentioned Google Scholar is open to everyone and quite good anyway so not sure of the issue?
Sounds more like propaganda.
But that doesn't solve the access: Finding the PDF you want on the publishers website can be a huge pain. So people just paste the DOI/Pubmed-ID etc. into Sci-Hub to get right to the PDF.
Hope that makes some sense :-)