mmm cc3 is a lot for a 2/3.
you play it turn2/3 and then wait one full turn to get your extra land. In mid game it's easier to get that land and play it. Without coven is too slow (even if intel is always good).
It pitches !
in the end it's playable but slow/unexciting.
The reason I have rather little hope for this card in vintage is the decks that can use that body effectively, get creatures in the yard fast enough to activate it, and make black mana to cast it, would much rather have cards in the yard for things like the gak then cast a 2 mana 4/4. This is a bad sign but the ability is interesting and maybe saves it? Unfortunately, no delver would really like its spells back in it's hand but hogaak wants them in the yard to take advantage of. Keep in mind I am not a zombie lord my skill lies far more with the doomsday decks so I may just be wrong but this one looks like a miss to me.
Gak requires far more of a commitment to include, since the delve number is so high. This needs 3 creatures which you can typically just get in the yard through course of play. There are probably also some shenanigans you can do with Anger and the like to make this more aggressive. It reminds me a little of Soulflayer, which has always been a pet card of mine, in that it needs a creature heavy shell to work, but as pointed out Wraith and Deaths shadow both work great here as well.
@botvinik There are a non 0 number of decks this affects that are not specifically blue decks. Looking at the deck by color as opposed to archetype in vintage is a lacking evaluation strategy to be honest.
Getting this down early against shops can yield a lot of damage on that first turn. Likewise survival decks cast a lot of free madness stuff that all trigger this.
My biggest question about this card is honestly is it actually better than pillar? Generally speaking when you play pillar you do not care about the self damage so this being one sided may not matter much in many matches, but pyro can potentially hit every spell, not just the ones after the first, and is not a targeted card. Pillar does not see much play so if this card is questionably better than pillar it would be hard to imagine it seeing much play.
This card is note worthy not because it’s particularly game breaking but it is the best “Terror” yet to be printed. Doom Blade has had a very minor roll in the sideboards of Dark Ritual decks since it can be used to kill a Lodestone Golem, Eldrazi, or random fish creatures such as Thalia, Magus of the Moon, or Collector Ouphe. Infernal Grasp clearly has more reach than Doom Blade now being able to hit DRS or Leovold. There may be better specific answers to things, but you’d be surprised at the lack of flexibility amongst black creature removal. This obviously isn’t breaking the meta, but it is note worthy and I will be picking up a few just to have on hand. I put it into the “Ingot Chewer/Goblin Cratermaker” list of cards you just ought to have on hand for some future meta you want to sideboard against. Terminate is a fringe modern staple that hits things beyond the scope of Unholy Heat or Prismatic Ending (Murktide). This basically better than Terminate, although regeneration could be relevant in the case of Welding Jar.
In a format with 4 gushes I think this could be fantastic, but as of right now it may just be to hard to pull off. It could be ok if the other things it brings to the table (no max handsize and repeatable draw engine) were needed, but I think it may fall short.
I don't think this card is comparable to Glimpse the Unthinkable, but rather to Sadistic Sacrement or Jester's Cap in that the matches you would side it in for are ones where you know that removing key cards can be backbreaking.
It's not going to reliably hit the key cards, though. let's say oath is our example. the oath deck has a good number of 3s and 5s in walkers, force of will, also probably a dig through time, in addition to its targets (let's say 2 griselbrands). if the card is only successful if it hits both griselbrands, it's a pretty bad card. that's only going to happen about 2.4% of the time, maybe a bit more or less depending on the list.