Rework RemoveCounterCost

[Susucre]

RemoveCounterCost is a cost for removing a number of (specific type) counters from permanents or cards.

Currently, it does take a non-specific Target. Notably there is no constraint/check that the target has at least one counter. Which results in an unintuitive UI that rollbacks the payment if selecting one with none of the counter.

It would make a lot more sense to provide Filter(s) instead of Target, and add the extra minimum constraint to the filter to target permanents/cards with at least 1 matching counter there.
It could also make sense to split the class for FilterCard (rare, e.g. [[Mari, the Killing Quill]]) and FilterPermanent (more common, e.g. [[Falco Spara]]).

Another solution (worst in my opinion), would be to always make sure that the target is filtering permanents/cards with at least 1 of the matching counter, but that seems like a lot of extra code in userrs that should be shared by the common effect.

[github-actions]

Mari, the Killing Quill - (Gatherer) (Scryfall) (EDHREC)

{1}{B}{B}
Legendary Creature — Vampire Assassin
3/2
Whenever a creature an opponent controls dies, exile it with a hit counter on it.
Assassins, Mercenaries, and Rogues you control have deathtouch and "Whenever this creature deals combat damage to a player, you may remove a hit counter from a card that player owns in exile. If you do, draw a card and create two Treasure tokens."

Falco Spara, Pactweaver - (Gatherer) (Scryfall) (EDHREC)

{1}{G}{W}{U}
Legendary Creature — Bird Demon
3/3
Flying, trample
Falco Spara, Pactweaver enters the battlefield with a shield counter on it.
You may look at the top card of your library any time.
You may cast spells from the top of your library by removing a counter from a creature you control in addition to paying their other costs.

[JayDi85]

Yes, it must be improved. Each related cost usage must be carefully reviewed -- if you put filter to current target (CounterType already has predicates) then it will fail due unsupported canPay method (see zero counters use case on X = 0):

P.S. Not related to issue -- looks like MariTheKillingQuill and OblivionSower uses wrongly face-up predicate (I don't see any requirements for it):