This is interesting. It seems there is a growing burden on consumers of information to understand the intent set up by website owners and technology maintainers: Aaron Schwartz died because his use of curl was deemed hacking. A judge here is making a distinction between linking and linking that bypasses a paywall.
I'm afraid that people will one day be required to divine whether or not a site was meant to be viewed by them: suppose page A displays links to a paywall-less version of page B when the user agent is "GoogleBot" in order to allow news to be indexed. Meanwhile, all other user agents see a link to a paywalled version of page B. Suppose a bug causes that page to display paywall-less links to all agents, and visitors to that page follow through to the non-paywall version of page B.
Under this interpretation, visitors to page B's non-paywall version could be exposed to allegations of unauthorized access to computer systems (or its EU equivalent) and be charged with hacking. What's worse, it seems the burden to prove that the link was shown to them in a legitimate fashion falls on the accused, and there's basically no way they can prove that.
I'm not a lawyer, so my interpretation could be way off, but this seems like a credible threat.