I'd support the last summarizing claim of the essay:
"As far as I can testify, there is currently no fertile ground in modern science to welcome possible new conceptual paradigms should this happen. There may already have been groundbreaking new ideas and they may have already fallen in to oblivion in some forgotten corner of the internet. Science may have already missed important opportunities for progress, and modern science cannot afford to lose even one of these opportunities."
It shows, together with the essay abstract, that arXiv problems (which I know from my own experience) are only particular manifestations of the underlying, much deeper science problems. After all, the main purpose of the arXiv project was open access to peer-reviewed papers, rather than open publication of any papers. The latter was rather our best hope, seemingly realized at the beginning but then fading away in arXiv practices.
In any case, there is an evident compromise solution for the arXiv situation, where the currently rejected or downgraded papers are still published in the originally (correctly) selected sections but with an editor's "stamp" or "warning" on them of something like "Author's Responsibility Only" or "Not Confirmed by Peer Review" (alternatively, there can be a special subsection of each topical section for them). And there is an easily accessible option for users to see, search or receive news alerts about all papers or only "officially reviewed" papers (without "warning") or only "bad guys" with the "warning", while otherwise all papers are easily and equally accessible. With this kind of solution, the system-science administration and official science lovers can be satisfied with this clear distinction and the possibility to neglect the "doubtful" content, while scientific freedom lovers will profit from equal accessibility of all papers (and the "unfair" stamps won't be practically important or can be even advantageous for respective attitudes).
And if this obvious solution is not applied to the arXiv operation, it means only that other open-access repositories, which do not want to use the open publication principle, can profit from this possibility. Although finally, open publication will win (and is already gradually winning), simply because every real possibility will find its way in a technically dense world. And also because even the most open publication and discussion won't change those deeper science problems, they can only provide more hope for the new progress in science...