I support Hacker News moderating itself however it chooses. However, if we are looking at it as a moderation model for large, open, non-editorial platforms (Youtube, Facebook) -- which I believe should all be covered under public accommodation law -- it clearly fails. And even if when we are looking at ostensibly neutral, publicly-orientated sites like newspaper comment boards, it fails.
Hacker News moderation is not appealable, not auditable, does not have bright line rules, and there are no due process rights. It simply does not respect individual rights.
So while this moderation method succeeds for Hacker News, and perhaps should become the model for small private sites, we should not try to scale it internet-size companies. Platform companies (Google, Facebook, Twitter) and backbone companies (ISPs, Cloudflare!) need a different set of rules geared towards protecting individual rights and freedoms instead of protecting a community.
Isn't every community -barring a few wild west's- like that? One Usenet, someone with an unpopular opinion would end up in people's kill file. On IRC, someone with an unpopular opinion would end up on people's ignore or would get kicked off the channel or klined off the server. Except for the kick and kline the effect of a kill file or ignore is akin to a global shadow ban.
The solution to the problem you mentioned is that unpopular opinions with merit will find their way to become eventually popular enough that they're adopted. Whereas unpopular opinions without merit eventually end up existing on the cesspits of the Internet. Because somewhere on the Internet, any person can spout their unpopular opinion. The question is, who reads it? Is it so much different from a shadow ban?