You still can't compare the ease of an access. It is like Google+click to read the article vs emailing authors for every article you want to read. At least that would decrease reading efficiency of the related research, at most - reduce the coverage of the studied material, just because it would require substantial amount of time, compared to direct click. That, in turn, might reduce the percentage of ideas that came during reading others' articles. Think of Sci-Hub as a hypertext engine for science. You copy DOI, you press the Enter, and you are reading an article.

Recently quite a few countries in the EU (including Germany), along with the University of California, have cancelled their contracts with Elsevier et al. France has not yet joined the rebellion.

Just speculation, but maybe this is a pre-emptive strike? Probably the courts would be less inclined to support the publishers if they sued to block these services only after they entered into conflict (with the universities)?

I think it's even simpler than that: the universities and research facilities don't use the three main commercial ISPs and won't have any kind of block.

Not true. Most people I know of or personally in academia in Germany and other countries rely on scihub. It's an incredibly valuable resource.

I'm a few years out of date, then, and I'm glad to hear that the situation has changed.

I would expect using scihub to be more convenient than any other option except direct google scholar PDF results. I have access to my uni's resources and journal subscriptions, yet I don't know how to use them (and frankly I don't see the point). It's extremely likely that circumventing a DNS block is less of a hassle than the procedure to access paywalled articles in a large majority of institutions.

Unpaywall is also a great method of finding alternative, freely-accessible versions of an article online: https://unpaywall.org/