army: still too weak! at that cost 6+gp it should be on pairs with KC or possession!
my approach: like thieve but for vp cards.
Army didn't have a GP cost. It was $6 only. Some people commented that it was too weak, though I personally think it's quite strong. It can always be a Lab with flexible filtering, which in itself is worth $6, but it adds additional options that make it very flexible.
It was pretty much the only card in the set that didn't have major flaws so I didn't really say much against Army, though I think there are some potential problems with it as well. First, it is probably fairly slow to resolve. You've got to look at a card, decide if you want to keep it, then choose a bonus if you didn't. Often you will draw a new card, in which case you have to evaluate how that fits in with your current hand. Then you have to look at
another card, decide if you want
that now, and if not you have to pick another bonus again. Spy is already considered slow to resolve with more than one opponent, and Army has a similar number of decisions to make. It could also be tough to track -- it's kind of like Pawn but with other events interrupting the choices (the filtering part of the action, which happens twice) so it may be tougher to remember whether you happened to choose a bonus other than +cards.
Those are nitpicks though and would have needed real play testing to see if they were actual problems. There were bigger issues at hand.