I think you're being way too optimistic here, if you think AI is some sort of magical ticket to produce good images (ones that actually match what the author wants, as opposed to AI getting a prompt and paining whatever, that has little to do with what the prompt asked for) in manner of minutes and whole updates worth of them in just a few days.I know you have seen what a couple in the FanArt thread can do on their own, and they aren't spending days doing it. They have been working on it for awhile at this point, and are getting very consistent images now. They can drop an image described to them in minutes. Compare to EK including 30 to 50 non-sprite images (a wild guess that seems close to me) in an update, I would guess a few of days of generating and tweaking images to suit would result in a lot more than that. They wouldn't even have to come up with the scene themselves, so less thinking more prompting.
You only need to take a look at some of the AI-based games (the ones that aren't just the latest NTR phone game cash grabs with a dozen images in 0.1 release and "please support my Patreon if you want more") and how it can still takes them multiple months to get an update out, that's nowhere near the level of complexity ORS gives, so it's not like they particularly struggle with scripting and writing. This is often combined with complaining/excuses how the generation is actually hard and won't people just stop bitching the next update was delayed for the third time, already.