(I haven’t read the contents of that paper btw, so this is just a general comment on moderation and moderation transparency.)
Discourse is already excellent and better than most forum software - users get notified of most mod actions. However I feel we could certainly do with expanding moderation features.
To give you an example, we routinely edit out bad language or personal attacks from posts and some users have complained about this - asking that they are asked to make the changes themselves and that we don’t touch their posts.
Of course that is not practical - the whole point of a moderator removing personal attacks is so that the person that the attack is aimed at doesn’t see it, which could lead to things spiralling. Additionally, moderators don’t have the time to PM a user asking them to make the change, make a note of the incident, remember to check, PM the user again or take other action if they haven’t made the required changes - it’s just too much of a burden on moderator teams.
An alternative could be to use the flag system which hides a post - but what happens if the user just unhides it or does not make an adequate change? We’d still have to make a note to check, unhide again, etc. (Additionally, removing the post from view also has other negatives - can slow the flow of a conversation, can make people feel like it’s too much bother posting, etc - which can all avoided usually by a very simple mod edit.)
I do firmly believe that Discourse could take moderation to the next level just like they have in other aspects of forum software 
It would require a lot more thought than just a single post or thread though - moderation is hard (nobody likes being moderated!)
I’d be happy to be part of such discussions and I am sure @HAWK would be invaluable in such an effort.