I feel like a project to make an adobe CS competitor, limited in scope to those early versions like CS1 or 2, could really get off the ground. I suspect for 99% of users, things added since the CS1/CS2 days aren't really used enough that not having them becomes a dealbreaker
Fully agree.
I think this is true of a lot of software, really. Very few people need more out of their office suite than what’s available in MS Office 2000 for example, which is somewhat validated in the sheer popularity of GSuite (with e.g. Docs being little more than WordPad/TextEdit with collaboration bolted on).
In general I also think that fixed-feature projects are a woefully underexplored space in the FOSS scene, which is a shame because they’re structurally more sustainable than projects that endlessly add new features.