I completely agree for those people associated with large universities who go to Sci-Hub it's a pretty strong statement that our systems are sewn together with the gaping loose stitches of a forensic pathologist after an autopsy. The job of links like "find text" is to take the user directly to the content, not to another screen that they have to figure out and even then might not work I'd like to see publishers recognize the IP ranges of our universities so that clicking on an article through Google Scholar takes them directly to the article. Yes, they'd need to know if we have access through an aggregator but it does no good to complain about theft of articles when our systems essentially punish users for attempting legal access.
We would do far better in academia and publishing to adjust our sails instead of cursing the wind. Or, as Patry put it, no law is going to save the business model of a company that refuses to give the customers what they want. (don't have my copy of How to Fix Copyright with me or I'd give a proper citation.