For traditional books, a huge reason why publishers get the lion’s share they do is because of the risk they take. The cost of creating the books themselves, all the logistics involved in getting them to readers (through bookstores or other retailers), and the advertising as well. There is a lot at risk for them if the book sells poorly, risks that the author themself doesn’t take.
There is no such risk for publishing a choicescript game. There’s a copyeditor, a small database hosting fee for user saves, and the time involved in going through all the app/game publishing menus on the app stores and Steam (which takes a matter of hours, not months of work). That’s not an extensive list of everything involved in the publishing process, but my point is that the monetary risk and time just isn’t comparable. Continual support through emails is definitely something, but it’s minimal risk and cost compared to a misprinted pallet of books.
The CoG/HG business model is one of minimal risk and very low overhead. The long and short of what they offer to authors is a marketplace and a built-in audience. I would argue that by this measure they function more as an app store than a publisher, so instead of a lion’s share, something more like 30% would be appropriate. Perhaps in a future when pigs can fly, we’ll get there!
I do foresee more authors going the self-publishing route in the future, though, and I think there’s a few reasons for this. Open-source tools like Twine are a great example, but even if you wanted to create your own text interpreter from scratch, it’s fewer and fewer chatGPT prompts away. The main problem has and always has been getting an audience, and that’s why I think we’ll see authors start with choicescript then continue their IF writing careers elsewhere.