It's a lot more complicated behind the scenes than it probably looks like up front. For example, one reason why a good-seeming article might be killed is if the site is banned because of past abuses. But for cases like that, vouching is good.
Beyond that I'd need to see specific links to say anything meaningful.
OK. Here are just three I spot at the moment from the first 2 or 3 pages I see, when browsing the site by the /new sorting:
1: [dead] Protect your S3 buckets from breaches (mklein.io) [0]
2: [dead] The NSA just released the preview version of Ghidra Debugger on GitHub (github.com/nationalsecurityagency) [1]
3: [dead] How to Write a Terminal Multiplexer with Rust, Async, and Actors – Part 2 (hashnode.dev) [2]
I'm not saying any of those is a fantastic article, or one that particularly interests me. But I can't see anything particularly 'wrong' with them in an abusive, spammy or 'irrelevant to HN's readership' sense. And, as I said before, there are usually at least one or two such submissions, which seem to have been killed for no apparent reason, on every page when browsing by '/new'.
Interesting you should say there's more complicated filtering going on behind the scenes because, from this side of the wizard's curtain, it does often look arbitrary, almost like people are flagging stories for personal reasons other than completely objective ones. And, as I said previously, one of the reasons I had turned 'showdead' on.
PS: I know I bang on about this all the time. But it would be nice if there was some kind of filtering on HN that blocked people endlessly posting the same stories over and over again. Here are today's entrants for "Keyboard Headbutter of the Week":
Lots of forum software will pre-scan what you are about to post for keywords and ask "Your topic seems similar to these <list of posts>. Are you sure you want to create a new topic?"
It would be good if HN would do the same.
Mind you, given that so many people obviously think that; 12 or more hours after a major HN-flavoured story broke it's not worth checking to see if anyone already submitted it, before submitting it themselves --they'd probably still headbutt that Submit button anyway.
Beyond that I'd need to see specific links to say anything meaningful.