$0.02 regarding the name of the feature, which I think anchors too much in the "censorship/free speech" principle discussion when it seems you have much more tactical goals (moderator workload, foster specific topics, etc.).
Consider renaming the feature from "pending comments" to "civility filter." In prose one would say Moderators now can require civility via special-purpose moderation on specific threads.
Similarly, consider changing the "[pending]" text to "[Pending civility check]" and changing the "endorse" link to "Affirm civility" (or just "Civil" to keep with one-word nav elements).
Consider renaming the feature from "pending comments" to "civility filter." In prose one would say Moderators now can require civility via special-purpose moderation on specific threads.
Similarly, consider changing the "[pending]" text to "[Pending civility check]" and changing the "endorse" link to "Affirm civility" (or just "Civil" to keep with one-word nav elements).
Why?
"All speech must be endorsed" --> "Bad Thing"
"All speech must be civil" --> "Good Thing"