The story development is good. What's here is elegantly presented and reads like a compelling and surprisingly believable relationship drama. For the sake of just creative momentum, I would probably urge the author to continue with writing a story that they were satisfied with first. Yes, it means ignoring the rest of my own rant. Don't focus on creating additional branches or populating other possibilities. Just deliver good or better scripts.
Afterwards, leverage the scaffolding of stats that you've created, and then add in possibilities.
The existence of the meta statistics were enough to get my interested into the choices and even into seeing what was happening under the hood. They are definitely a smart inclusion. I do agree with some previous comments where the BJ in plot 8 seems forced. There's plenty of outs where an NTR-optional set of activities could be insinuated instead if the stats are leveraged.
More logic leveraging the statistics would be great. There's a few ways statistics could be used to mitigate or direct different results. Arousal becomes a feature that builds up over time when the player makes any decision, and once it reaches a particular point, she decides then to commit to an activity.
With high corruption, Lucy could indulge her urges in a publishing to an OFs. Low bond she commits to cheating. High arousal, but no relief at the end of the plot might result in a reward of a "sexting" session, justifying an arousal reset when going into a future plot, keeping things almost episodic. Given the current framework, one could probably plan for 3 possible outcomes during any given episode, all the while avoiding branching story complexity.
I, generally, hate renpy vn's. But this one is rather good as it is so far. So, doubly good on you