Also a testament to the sustainability of usefulness of having a business model that isn't directly about monetization, which means it's probably really hard to copy.
HN is a PR project for an accelerator/venture capital firm. Since its purpose is to make ycombinator look awesome and generate leads for them (providing indirect monetary value), we all get a fun place to hang out and discuss stuff so they can project that image, but I have no doubt that without that fairly unique situation it would quickly devolve into some variation of what we commonly see with social networks that have to find a way to pay for the services they provide.
The technical side of HN isn't insurmountable, the moderation side is harder but the community side is the hard part.
I could build a HN, I could maybe moderate it (though probably not as well as dang and co, they are rather more even handed than me) but getting people to come and use it without attaching it to a VC fund not so much.
If I wanted to make a profit though then things are different.
Good community projects can self fund, lichess is a good example, hundreds of thousands of users, millions of games a day and zero advertisings, scheezy tactics.
Completely funded on donations, even pays the lead developer a liveable salary (though certainly not what he'd earn on the open market working for said scheezy social networks sadly).
I actually wasn't trying to make a point that HN itself is specifically hard to copy, but the more generic idea that having a business model that's not based on monetization is hard, for some of the reasons you noted. I guess a corollary is that HN is therefore hard to copy, but it's not for any real technological issue, as you note.
I also agree it's not impossible, as you point our with lichess, it's just that with the current level of consumer awareness of privacy, I think most people (still) don't realize what they are paying for some competitors that appear free, but are really just monetized through selling personal information, and for some services network effects trump almost all others. A free chess matching service, where being linked with random people of appropriate skill level is a selling point is a lot different than a social network that's built around connecting with friends and family.
AS to being funded on donations, while I think it's wonderful that a project can be funded on donations, I wouldn't want any project I care about to be in this situation unless the donations exceed all operating/staffing costs by a very comfortable margin. There's just not enough leeway for unforeseen problems in a situation like that. Server failure? Severe illness in key personnel? Massive influx of new users that haven't matured in the user lifecycle to consider donating yet? That's such a stressful situation that I can't imagine wanting to live it for more than a few years. Unfortunately, if you can't get enough donations to get past that point (or keep increasing scope until you are at that point no matter the incoming funds), I'm not sure a solution besides monetization in some manner, and the problems and perverse incentives that come along with it. :/
HN is a PR project for an accelerator/venture capital firm. Since its purpose is to make ycombinator look awesome and generate leads for them (providing indirect monetary value), we all get a fun place to hang out and discuss stuff so they can project that image, but I have no doubt that without that fairly unique situation it would quickly devolve into some variation of what we commonly see with social networks that have to find a way to pay for the services they provide.