Thanks for responding, I’m glad my comment gave you some things to think about!
Mu Cartographer is a very cool example, I think part of what makes it work (at least from a distance, having not played it) is the visual spectacle of the generated landscape. Very much feels like someone was playing with landscape generation and thought “How can I make this game?”. I think it adds credence to the question “Would exploration feel engaging just from interesting visuals?”. I’m a huge fan of Yume Nikki and that type of exploration and I think there’s very enjoyable on it’s own to be in a world of the wondrous and unexpected. There is a bit of concern with same rooms appearing too often and loosing their novelty, but I think that could be solved by giving some rooms small pattern requirements (Valley room next to two desert rooms or something like that) so they have “identity” and are a bit more rare, as well as some rooms that are just rare in general. I think ideally with the right amount of variety, repeated rooms could become comfortably familiar, instead of tired.
My gut does still say there could be something more, some way to get that “balance”… It’s a tricky question and might just require some trial and error prototyping (unfortunately time consuming!). Maybe like Mu Cartographer, you could have some sort of sensor that tells you how close you are to completing the pattern. Could give that fun sensation of: “Agh, I’m so close, but where is it!”. And yeah, I like the idea of giving some flexibility on meeting the “correct” solution, would definitely take some strain off of repeating attempts!
Anyway, I love to read about designers’ thought processes so posts like these are very much appreciated. It’s a huge challenge to ‘map’ out (haha) relatively unexplored areas of design, so its really interesting to see how people grapple with those unanswered questions and think about all the possibilities. I wish you luck working on this idea and others!