Thanks for providing some food for thought on the specifics of the proposal. These are real and justified drawbacks. My interest is more in access to the flagging behavior in aggregate rather than campaigns against any specific person. But this being the internet it certainly creates a new avenue for deranged people to kick off. However does the risk of this hypothetical hate campaign:
a) exceed the likelihood of people doing this via commenting anyway
b) justify the opaque and powerful nature of flagging as-is
Perhaps you stopping flagging if you're not willing to justify a flag is a good outcome in aggregate? We have mods to kill threads which violate the guidelines already. But looking at the /active list there's certainly an amount of (probably organic) censorship of controversial threads in either direction (though my gut feel is it biases more towards censorship of articles about the latest outrages of US government).
I'm not really interested in say, Ruby, I think people should probably use languages which are type-safe if they want to avoid catastrophes in production and 1am pager calls. However if I see an article about Ruby I'm just going to not engage with it. Perhaps your existing interpretation of the unwritten rules is too broad and actually we ought to rein in the amount of flagging anyway?