First, Chess-bot research has been decades in the making and there's only a few pieces. Cardgames have far more interactions.
Second, players in pve campaigns generally prefer to be smarter than their opponents and win via superior play against a foe that has more resources,. Entire series like the fire emblem series relying on players exploiting the AI's mistakes and predictable actions. Players don't necessarily enjoy a game that's outplaying them and making them feel dumb. They mostly want the AI to avoid any super-obviously-stupid plays like using a removal spell on its own dude or failing to block lethal.
So it's a huge amount of effort to make truly great AI for a cardgame and many players might even prefer you not do that in the first place. When it comes to opportunity cost, would you rather hire 3 extra AI engineers to make the AI harder over many months, or would you rather hire 3 more designers/gameplay-engineers to make more adventures, champions, events, etc?