I know that topic and settings it is referring. In this topic is mentioned how to block login using robots.txt. But robots.txt doesn’t block much, because following, or even looking it, is a choice.
And I’m still confused because that one sentence:
Does not apply if allowlist is defined
If allowlist works totally opposite way and is allowing only whitelisted user agents and blocking everything else that would be so dangerous setting it should not be there at all.
There is really wide variety of user agent, and I’m quite sure Discourse can’t do things such way it can detect when an user is a human and when it is a bot, and after that decide when it uses whtelisting and when not.
That means too that if there is real blocking and someone decides use whitelisting that poor admin will close doors to everyone just because only googlebot is allowed. And trying whitelist every ”human” user-agents is… just impossible.
Every topic that I found tell how to use block-setting, not what it does. Big difference.
That is why I would like to see somekind codex here that digs in bit deeper that short helps. And sorry, GitHub is not the right answer. It is wonderful place if one would like to do dev-job. Otherwise and for admins… no.
So — I still don’t know what blocking and allowing means here.
And I’m totally sure this branch is not anymore how to find alternative to adsense 
But I don’t flag myself anymore just to ask moving to another topic — I’ve learned my lessons.