depends on when the Steam version comes out, if it comes out in the spring then a Chapter 25 in the summer/fall would be conceivable.
otherwise not before winter, as you say
Yes, I have no knowledge on latest updates on the progress in the roadmap of the redo for Steam.
I guessed that the redo would have taken approximately six months just based on the update from Patreon that was shared here soon after the release of Ep24.
If I remember correctly, the Dev stated that:
1) the redo was involving (at least) the first 10 episodes (or maybe even 12, I don't remember exactly, but anyway not just 3 or 4 episodes).
2) regarding the art it wasn't just a matter of batch substituting old models with new ones and just waiting for render time of the whole batch, but a lot of manual tweaking was going on both in customizing the models of the characters (fine details like the shade of brown of the eyes) and in choosing among different models/parameters for lighting.
This process is more time consuming because of course he has to try different parameters for the same scene, then launch four or eight renderers of it, wait for them to be completed, compare them, and decide which parameters give the best results before being able to set the new parameters to the whole batch...
...this is obviously more time consuming than just re-render the whole batch using the same source files but with just increased quality due to more powerful hardware would have been.
3) he wanted to change something in early choices/branches to improve game experience.
This is time consuming because it is error prone and requires concentration while coding and thorough testing to be sure not to break compatibility with subsequent episodes.
4) he wanted to refactor the game logic by collapsing two state variables for each character into one.
In my limited experience, even the most simple form of refactoring, that is changing just the name of a variable is nontrivial for a program bigger than a few hundreds lines of code if the IDE doesn't provide integrated refactoring facilities like Visual Studio does, and as far as I know Ren'Py doesn't provide anything like that (if I'm wrong about that, corrections are appreciated).
Besides collapsing two variables into one change the logic of the program and so this kind of refactoring can't be automatically done, but instead manually he has to find every occurrence of the two variables in updates due to choices and in if clauses and substitute those with the new logic... I'm confident I'm not wrong if I say that's nontrivial to be done for 10 episodes while making sure not breaking the game by making some branches unreachable by mistake...
So I guessed six months would have been a reasonable time for the redo to be as tested and polished to be publishable on Steam.
---
Of course, if some roadmap updates have been published and I haven't seen them, my guessing can be totally out of date and so unreliable
