I agree with Stalker0. And as for how it scales, well, let's wait to see how it scales (or doesn't). I don't see anything wrong with some characters having a power that lets them do small amounts of damage on a miss. Yes, it means they can't fail to kill a kobold, but why is that a problem? It would make me tempted as a player to improvise "I attack two kobolds at once" ("Okay GM, can I pick up one kobold and beat another to death with it?") for variety, especially if there are lots of them, but I don't know that I'd complain about an autokill or three.
As for MM... maybe. The question is really how it scales compared to other people's damage. If the slayer fighter is doing expected damage per round of, I dunno, 14 on a hit /2 (because of missing half the time) plus 3/2 (because of missing and slayering half the time)=8.5/round, and a rogue is doing 7.5/2 (missing half the time) plus 3.5/4 (sneak attack half the time, missing half the time)=roughly 4.65/round, then the wizard's 3.5 per round at 1st level seems not at all problematic. (I worry if the rogue is then a tad underpowered, actually.) The wizard is lagging behind the damage curve, but that's correct, because the per day spells surge way above the damage curve. The question is, is it too far behind the curve? And there I'm not sure--the wizard can autokill kobolds, but can't kill an orc or hobgoblin even with max damage, even in 2 rounds, at the same time as the fighter is frequently one-shotting orcs and hobgoblins. But that might still be right because of when the wizard breaks out the sleep or does 8 hp on average to a group of 6 goblins that fail their saves against burning hands, killing them all.
The other question is how at 9th level, 14 points/round from the wizard's magic missiles will compare to the expected damage from the fighter or the rogue. We can't calculate that without actually seeing what the 9th level fighters and mages look like... but I predict that 14 points at 9th level will be no worse than 3.5 at 1st level.
I'm entirely willing to be proved wrong by play, but at an analysis level, these don't seem like big problems.