I'm not sure they have fully figured how to deal with Reactions for instance, and all those similar 5E rules that give a player decision points when it's not their turn (the Diviner mage being just one example that works the same way). And I feel they need to rethink some core aspects of combat. So, at the same time, they'll want to know how combat works to know how to translate a 5E Class into a BG3 Class. At the same time, they'll want to have a first draft of all classes to have a better view of how combat works.
Good points. I have no idea how they intend to go about reactions (and whether they intend to do anything about them, tbh) and I don't envy them having to think of something. It's a system that's very hard to translate into the video game format. I've seen a good post on reddit on reactions, where the OP explained WHY they're so problematic and what are the problems with different approaches. Yikes. I think the best option would be to optionally be able to script when they trigger, but it's not a perfect solution.
I wonder about the classes, I'm fairly sure they have much more than drafts by this point. But some classes have "probematic" or "labour-intensive" mechanics. I think the initial choice of classes was a mixture of "basic" (Fighter, Rogue, Wizard, Cleric) and something that needs a lot of time testing. Ranger, obviously, because of the rework. Warlock because of... potentially wonky mechanics? Alternative magic system (Pact Magic), pact boons, invocations... Lots of things to go wrong, and I believe we already see some of them not working properly. There's also the matter of companions - we only have those of the implemented classes, though the classes probably took precedence. (I'm curious though, because the class choice seems like it had little wiggle room, yet they've also said they've chosen specifically the "evil" companions, suggesting that it wasn't class-related. Hmm.)
Let's think about the others:
Barbarian: pretty straight-forward; I don't think rage mechanics are particularly difficult to implement. Probably left out because it wasn't a priority.
Bard: Bardic Inspiration has the same problem as reactions. I could see them implemented after they figure out reactions and similar mechanics, as you say.
Druid: wildshape. I could see it as problematic for several reasons. Animations, animal choice approach (all of them? only the ones you've seen in the game?), inventory management, interactions...
Monk: ki. I don't think it's particularly hard to implement, but it's a new magic system to code.
Paladin: smite - similar problem to reactions, but I could see them just going with "you decide beforehand".
Sorcerer: metamagic and sorcery points, two new systems (or one complex system) to develop. Some of metamagic options might more problematic than others.
So if they were to introduce the remaining classes in the order of low- to high-hanging fruit, I'd say it would be: Barbarian, Paladin, Monk, Bard, Sorcerer, Druid. More or less, especially depending on how careful they want to be about "invoked" mechanics like reactions, Bardic Inspiration and smite.
