[goatkarma]

Err I read TFA...Just read the top comments on this HN post to see ease of use is absolutely key to why folk use it.

My perspective is from the University side of things so can only speak for that rather than from the perspective of users who do not have access to the content at all.

I've spoken to users who workflow consists of googling the article title to get the DOI, then putting that DOI into Scihub to get the PDF, without even going near the University's library system. Most of the time the Library actually has an electronic copy, but the process to get them, even with SSO, is laborious and confusing. Just look at the SSO login screens for different publishers sites : some say 'sign in with single sign on', others say 'institutional login', or 'Shibboleth login' etc. How are University users expected to jump through these hoops when the can search for the title or DOI and get the PDF instantly?

[cjslep]

And my perspective is from someone who does research on small-world networks in their spare time.

For me it's a matter of having access. I can easily find seminal papers like Kleinberg or Watts-Strogatz. Anything that is more recent and builds off of these papers tends to be locked up and I'm not able to afford buying these papers one-off, since I don't even know for sure they'll be in a direction useful to my research.