What if the "lose the game clause" was kept, but was something like getting Hellbent or Threshold (a little bit easier) or some other gamestate? If the card is ridiculously broken and beyond repair, then I have no qualms about this being locked and going back to the drawing board.
The "lose the game" clause doesn't mean anything. Nobody's going to find a way not to shuffle their library. Instead, they're going to find a way to draw a card. As you suggested, they'll try to use this frequently with Brainstorm so that they can not only get the card they want but they can shuffle away two cards that they don't want.
I don't really have my finger on the pulse of any format where tutors are potent anymore, so I will defer to Anusien's judgment regarding this card. If he says that it ought to cost
and be a sorcery, I'll trust that he knows what he's saying.
I don't think that you need to lock this up and go back to the drawing board, though. The concept is sound even if the execution is flawed. I think you may be hung up on the idea that a "pact" card needs to cost

. I think that design space is razor thin. On the other hand, if you're open to developing this card, like we're trying to do, I think that it has good potential.