Any chance of seeing what happens if the needle is moved the other direction? That is, see what happens if you actively discourage people from flagging stories simply because they are political, and reserve flagging for comments that are incivil and abusive?

I think user 'opsiprogram' describes it well here: https://news.ycombinator.com/item?id=13111900. It's unlikely that the outcome will be ideal, but like this experiment I think there might be useful lessons that can be learned in a short time.