I agree that writing new code this close to the hardfork implies risks. And leaving it to the next iteration is probably the safest way to go considering the moment in time we are in. But it should probably be addressed in the next iteration, it makes little sense to mix witness voting with SPS voting.