You'd be welcome to email specific examples that you're worried about to hn@ycombinator.com, but based on everything I've seen, the null hypothesis has to be that your perceptions are conditioned by your own passions about said "certain entities", whatever they are. I don't mean to be dismissive (and when people ask us to look into cases they're worried about, we always do); it's just that this is by far the most consistent phenomenon I've observed on HN. Nothing else comes close.
Internet users invest far too much energy into perceptions of "asymmetric astroturfing" and other sinister manipulations. Most of what they think they're seeing is easily shown to be a figment of perceptual bias, driven by their own passions on the topic. When I say "easily shown", I mean even from the public data. People could, but generally don't bother, to check the history of the accounts they accuse. When an HN user was posting about, I don't know, Ansible in 2018, or Forth in 2014, what are the odds that they're a foreign asset or secret shill on $hot-topic of 2021? That's literally what most of this stuff amounts to.
What most needs to happen is for us all to become aware of how these psychological mechanisms function in ourselves, so we can interrupt the process by which they convert into sinister fantasies about others. This isn't unrelated to the "balance of the community at large"—it's at the heart of that question. Not seeing, or not wanting to see, the genuine variation of the community on divisive topics is what leads to sinister fantasies. The odds are overwhelming that the person against you on $hot-topic is not $evil-manipulator, they're simply someone with a different background than you, who is as sincerely persuaded as you are.
Does genuine manipulation also exist? Yes, but even answering 'yes' to that already gives a mistaken impression of the proportions. The group-psychology dynamic I'm describing is responsible for the overwhelming majority of these perceptions. That's why the site guidelines ask people not to go there in the comments, but rather to email hn@ycombinator.com if they have concerns. Then we can look at the data and see if anything objective can be found.
I hope I haven't bitten your head off! My comments on this can get tetchy because of the number of times I've had to repeat myself on this issue. At some point I'm going to compile all this into maybe a single explanation that can just be linked to.