Whilst I am pleased to hear about a card game going in a new direction, I can’t help but feel a bit leery of hearing “… base set consisting of about 200 cards…” and then “I also have a terrible problem with not finishing things I start…” in the next breath. I don’t wish to disparage you or your ambitions here, but to avoid setting yourself up for a fall, overwhelmed with the prospect of balancing so many cards over so many characters (and that’s not even considering the other aspects of the game as a whole) a more planned approach should be in order.
Before running away with character and card designs, you’re going to want to focus on a proof of concept and for that you’re going to want to pare things down - like a lot. Firstly though, a quick comment on the core gameplay loop (ie, the moment-to-moment actions the players will be taking in your game) Thankfully the loop initially sounds solid - pick teams of characters and battle them - fair enough, so would I be correct in inferring that there will be two general phases of the gameplay loop, essentially “combat” and “out-of-combat”, the latter of which I assume would cater to deck management? If you envision the player doing more you will need to define this first and foremost. For now I will just work on these given assumptions.
So let’s start simply. In almost all deckbuilding card games, you have a set deck of starting cards. Slay the Spire, for example, boils down the 100+ card options to just 3-4: strike, defend and 1-2 signature cards. Other card games will have more complex setups that affect your starter deck, such as gear/equipment loadouts, but you are still ultimately presenting the player with a starter deck that not only a) keeps their starting decisions simple but also b) still manages to convey a style of play that is unique to that character, class etc.
It’s those 1-2 signature cards in Slay the Spire that define each character, and those card choices aren’t an accident or an afterthought. It’s taking the broad overture of the playstyles or strategies available to a given character and working backwards from that endpoint so as to give the player the starting means to intuit and pursue those strategic goals when it comes to deck building - I’m assuming in the “non-combat” phase of your gameplay loop there would be a focus on deck building or management.
Your proof of concept should start small. Take a single team of three, or heck even a single character. What cards would constitute their starting decks? Using this initial constraint will help focus on the basic functionality of combat (ie, what’s going on in a battle) and start guiding your hand in not only defining gameplay aspects, the characters etc but also give you a better foundation for when opening the floodgates for new card concepts and strategies. On top of this, have a consideration for the aforementioned “out-of-combat” phase. How do you see deck management working?
If you are able to construct and demonstrate a completely gameplay loop, showing the phases both in and out of combat, then that will give you a great foundation to start building the game up from there.