Better leveraging Twitter's reporting feature is probably the most neutral way to solve this.
When a tweet is deemed response-worthy, they should post the report numbers. Value in numbers shields them in many ways and could legitimize their actions as a neutral party. Then, if they miss something, they can simply say there weren't enough reports. This will then empower the feature in the future.
I suggest this as active Reddit moderator with a community of 40,000+ subscribers who regularly has to enforce rules and uses auto-mod to help manage reports and shares that with the community.
-----------------------
You can report tweets for:
(1) Being not interested in it (you just get redirected to a mute or block button)
(2)It's suspicious or spam
---> The account is fake
---> Includes a link to a potentially harmful or phishing site
---> Hashtags are unrelated
---> Uses the reply function to spam
---> Something else
(3) It's abusive or harmful
---> It's disrespectful
---> Includes private information
---> Includes targeted harassment
---> It directs hate against a protected category (eg race, religions, gender, orientation, disability)
---> Threatening violence
---> They're encouraging self-harm or suicide
(4) It's misleading about politics or civic events
---> It has false information about how to vote
---> It intends to suppress or intimidate someone from voting
---> It misrepresents it's affiliation or impersonates an official
(5) It expresses intentions of self-harm or suicide.
-----------------------
It's pretty good but I would suggest the very simple following updates:
- Updating the main issue (It's abusive or harmful) to (It's abusive or encourages violence or destruction of property)
- Adding a sub-issue to (It's misleading about politics or civic events) with (A political official is supporting false or unsubstantiated information as definitive truth.)
- Adding a sub-issue to (It's suspicious, spam, or false) with (It's supporting false or unsubstantiated information as definitive truth.)
- Adding chevron icons (>) as a visual cue that each main reporting issue has many sub-issues