Overall, I like the idea for this project. I think the concept as it is presented would make for a fun and engaging game. I especially like the way that time has been broken down into a resource that is used to perform moves, and the way that this is represented to the player.
I think it would be interesting to see this game in action. It took me a few reads to put together a good idea of what you were trying to explain in your document regarding how time works and how character movement would work (granted I am a bit tired lol). But I think that would become a lot more clear in an actual game setting when the player can see what's going on, as long as some amount of tutorial exists to explain the different moving parts to players.
I have a few concerns regarding the scope of the project. You've divided most of the content into 'things-necessary-for-the-prototype' and 'things-necessary-for-the-release' well enough. I'm mostly concerned about how you plan on using reinforcement learning to train the AI to pick moves. This seems like a fairly involved process that might take a while to implement if you're unfamiliar with how to do it. If this is something you've done before, and know how to implement quickly, and you can estimate roughly how well the AI will turn out, then be my guest. Otherwise, you may benefit from taking some time to look into how this sort of training is generally done in whichever program you're using, and then deciding whether you want to continue down that route, or perhaps instead implement some sort of weighted decision-making algorithm that grades cards based on what the AI thinks it needs to do, and then picks the best choice.
Mainly I would recommend estimating how long you think it would take you to implement the different systems necessary to get this game up and running. If it turns out some of the systems don't totally work by the end of the prototype phase, I wouldn't be too worried. I would also estimate how long you think it will take to draw the art; you could do that by simply drawing one card or UI element, and extrapolating the time to the number of parts you need to make. Ultimately, I think this is for the most part a very doable project.
In regards to the "Outstanding Questions" in your document: you may want to consider first implementing forward-backward-only movement for the prototype portion of the jam (since that seems more straightforward), and then seeing whether a move to an expanded 2D grid would be a good fit for your project later, or if it's one you have time to implement. For that decision, I would take into consideration the way in which you implement attack range. If your implementation would allow a character to hit their opponent directly behind them, and you want to change that but can't do that in an easy way, then you may just want to stay with a linear system. If you're up to the challenge of figuring out all those shenanigans though, it could be pretty fun and great practice to problem-solve.
I hope the project goes well!