This is one of those setups that's clearly blatantly scumsided
unless
you break it, so the main balance consideration is in just how much advantage you get from the break.
There are (12×11×10)÷(3×2×1) possible scumteams in this game, that's 220. Assuming that the D:1 lynch is on town (if it's on scum town should have no real problem winning from there), there will be (11×10×9)÷(3×2×1) or 165 possible scumteams going into N1. The IC can choose between 8 different VTs N1; therefore, we can organize the possible scumteams into 8 groups of (165÷8) = 21, indicating the group via the choice of Innocent Child, and at the start of D2 there will be only 21 possible scumteams remaining. (The actual Innocent Child advantage via which the chosen player can confirm themself as town is irrelevant here, as it must necessarily duplicate the information transmitted by the original IC via the choice; the original IC can't choose a scum player to be the backup IC.)
As an attempt to visualise how useful this information is, let's consider a possible strategy for the IC: out of the 11 remaining players, there must somewhere in the player list (cycling around the end if necessary) be a sequence of a Mafia player, followed by three VTs in a row (there are 8 VTs left and not enough Goons to separate them into four groups of two). One example of a possible plan would be "find such a situation in the player list, then choose the first of the three consecutive VTs as the backup Innocent Child". Town will be going into day 2 with a guilty and three innocents; thus at the start of day 3, town will have two confirmed innocents and 5 VTs, against scum's two Goons. That's equivalent to a 5:2 vanilla game with N1 and N2 skipped; especially given that the scum kills are effectively all forced, and yet town has been free to lynch scummy players and VIs, I'd consider this to be townsided in practice (despite the EV being below 50%), as town normally do very well when scum can't get rid of their key players.
However, the strategy considered there is a bit suboptimal, as it only narrows down the number of possible scumteams to (8×7)÷(2×1), or 28, whereas 21 is possible. That said, the difference between 28 and 21 doesn't seem so huge that I'd expect town to have a pure break on this; they can probably make the setup clearly townsided after a D:1 town lynch (and outright break it after a D:1 scum lynch), but not more than that. "Clearly townsided but not 100% broken" is probably not balanced enough to run, though (assuming that players would even enjoy playing a setup that worked like this).
My first thoughts about trying to balance this would be to remove one VT and the D:1 lynch. D:1 would therefore become a discussion period where players tried to suggest plans for N:1, tried to arrange plans so that scummier players ended up in scummier slots, etc., but that ends with no result once people are ready. That makes the control-of-deaths more balanced, too (as neither side has an influence on who dies until after the IC's supercop information becomes public).
