I don't like the mob thing either but it's how large group dynamics on the internet work (by default). We try to mitigate it where we can but there's not a lot of knowledge about how to do that.
This could go after the reply button, before it, before the comment box even. Or, to get fancy, positioned based on karma (a little more in the way when it's low, more out of the way as it climbs).
No functional change.
I've found myself delete a lot of less useful comments when I stop and say "is this really helpful? Or am I trying to 'win' a discussion".
I don't have any info on whether it's mostly new accounts or randomly anyone that is likely to post rule breaking stuff so I'm aware I'm guessing on a lot of this.
Being able to turn it on for a particular thing makes sense (e.g. any political story) - my first thought on top of your suggestions is
* Show if under X karma
* Show if discussion has more than N flagged comments
Things seem to work pretty well here, hopefully there's some tweaks that don't change that but lower the burden for you and others.