The essential problem is. What lots of people have expressed interest in is Alexia using magic to gain agency. Which is, to say, a whole ass variable that needs to be accounted for every single scene in the game. Reviewing literally hundreds of Alexia scenes to add magic variables. And it would be a half assed attempt too, since to really deliver on the fantasy it would need to produce major variants in the events where Alexia's use of magic shifted the story's movement somehow.
And I can definitly see the appeal of adding in a variable like that, in terms of giving the player the chance to shape the story in a way they really enjoy. I can also see that to do it right, you'd need untold tens of thousands of words. perhaps even more. A massive development sink for a core staff member (someone who knows the material inside and out) for a large chunk of time.
Can't do it that way. Too hard. And if we did it without doing that, then half the logic of the scenes she's in wouldn't make sense or else her magical abilities would feel pointless or unaffecting.
Basically, I'm as down as anyone else for a magic Alexia plotline. But, if we did it, we'd need to do it right. And we just can't do it right with how things are with act 1 ATM.
Edit: As an aside, this is another area where you have to consider exponential growth in variables. That is, without any "bottlenecking" adding mutually exclusive variables with multiply the total content. But, adding non-mutually exclusive variables expontentially increasing total content.
Consider, right now, Alexia can be Rowan-Loyal, Jezera-Loyal, or Andras Loyal. This produces three content variants which if given their own entire segment, would triple the per-scene word count. But, Alexia doesn't just have that variable axis. She also has Low, Medium, High Corruption. Further, many scenes with Alexia feature Rowan.
So many scenes with Alexia already look like this:
Andras -A!High- R!Low
Andras - A!High - R!-Medium
...etc...etc..etc...
For a total of 27 possible Rowan x Alexia relationship states.
Now, in practice we dont' write remotely close to that much, because we don't give each variable entirely seprate text. Instead, we pin-point where the text can merge and "bottleneck" it back together. Theoretically, you can bottleneck as aggresively as you want. But, the more aggresive the bottlenecking the less substantial a variable feels. So as a writer in interactive fiction you need to be balancing the need for substantial variables (which inherently increase player engagement) vs the need to not die by workload.
Half the problem with the original Alexia Library Magic plan was that it would be done as a non-mutually exclusive variable. So, take the 27 possible Rowan x Alexia states above. Now, *double* it based on if Alexia is magic or not. You wind up with a whopping 54 substantial variables then and there. With more secondary variables like Feudalism, Alexia's Job Location, and Rowan relationship with side cast members that might produce smaller variables.
But, imagine for a second that we just ignored that and we didn't include ANY of those variables for magic Alexia. Well, we're in trouble. Because is Pure Alexia and Corrupt Alexia never going to use magic in diffrent situations? Are we not including magic in Andras/Jezera variable segments or their personal scenes? What about Rowan? Is Rowan going to respond the same way to magic Alexia in every version of it?
So it's not just that Alexia is in a ton of scenes. She is. It's also that she she's easily one of the HARDEST charachters to add a non-mutually exclusive variable at this point because all of her scenes are already so stuffed with variables that it makes Rowan x Alexia scenes into something of a meme among the staff as "the hardest shit to write".
