Earlier quoted context omitted.
I'm an academic. I've never met one who has objected to Sci-hub. Many of us use it, even to get papers we do have legit access to: journal publishers' access control systems are a pain in the arse. Leaving aside the societal merits, our incentives are to get published, then get read. Sci-hub helps with the latter, and doesn't harm anyone except the publishers (who most academics cordially dislike anyway).
I imagine even the most socialist/communist/whatever is at least partly incentivised by earning a living too. In fact, isn't that the problem, that quantity of publications (and citations) is used as a metric for continued employment? Those awarding research grants, etc., don't provide a free open-access publication as the place publication is required. It seems governments could fix this in favour of the public quit…
Governments and other major research funders are indeed beginning to do what you recommend and mandate various kinds of open access terms on publications. There are multiple options, some of which are reasonably favourable to the incumbent publishers (e.g. delayed OA release with an embargo) and some of which are not (e.g. it goes straight into our public database of funded research on day one). What the end game looks like isn't exactly certain at the moment.