By my lazy count, there's around 67 cards (including fetches) in my cube that interact, to some degree of favourability, with Brainstorm. I don't know if 20 copies per cube is correct per se and obviously not all 67 cards I counted are like, slam-dunk-always-relevant scenarios, but I think it's worth the slots I've given it for now, at least, even if all it did was help miracles (and it does more than that). It's a card that outperforms its competitors when appropriately tuned, and that sort of "kinda worse sometimes and clearly awesome other times" potential seems worth it to me. If people wanna throw it into a no-fetch/no-synergy deck I figure it's still drawing you a card a turn faster and might help in some rare corner case, so whatever. I don't really care if people aren't optimizing for it, because they usually don't notice or care themselves, anyway, and the potential added depth it provides for the occasional surprise corner use is worth it to me.