Oooh I made something similar to this a while back (link below). The score payout for this is interesting because if anything—this is a /townsided/ version of BV. As the game sample size increases, the town-scum win ratio should become more balanced.
The setup has no meaningful mechanical difference from the original BV. It has some built in failsafes against suboptimal play that BV is lacking, for instance lynching a guilty Day 2 after a nilla flip Day 1, but other than that the only benefit to suspect is reads. There are no bonus clears or chances at autowin that are not present in the original BV, but these can be made more immediately apparent.