I am tempted to say, "Hell no, we don't need that!" because I of course never troll or make toxic comments (who, me? Surely not, lol!) but that's a shallow reason.
Deeper would be that Steemit already incentivizes good behaviour so well that it scarcely seems necessary to put such a hoop to jump through for all participants. There are also a few questions that would need to be answered: how do you implement it? Does the post creator draft the questions manually -- or does a bot do machine-learning to the text and come up with them? Would it apply to comments replying other comments or only to the main body of the text (because I can see THAT gaping loophole for trolls and spammers -- and they sure as hell will too :D)
I would think it would apply to first level comment directly to the post. But it could be used to filter spam/hate on comments too. Pass it once per post and you can comment on anything there. The creator being responsible to filter is best.