Thank you! While the extra cards are just a side effect of very straightforward level generation, I had similar things is mind when making it. The game is inspired by the brilliant Mosa Lina, which I love because it's about playing with the system itself and not with a designer. I tried to replicate that in a pure puzzle game.
Ah, mosa lina, yeah, I loved that. It's been argued that it's an immersive sim. If so, for me it's the first immersive sim I liked. What it does so well is it maintains a good balance of difficulty, lets the player fail, and lets them quickly move on from impossible challenges.
If we were to talk about building a BLAME!-inspired immersive sim (also procedural, inevitably), I think finding ways to let the player fail would be a very salient theme.
The challenge of making large procedural worlds in interactive media is that it's sort of impossible to meaningfully interact with something millions of times larger than the player character (or having so much more information than the player could process in 20 hours), but it's not totally impossible, the player gets to make recurring choices as to which direction to trek in, and how to influence the weather movement of the constructors, but if these choices are real, this means the player's failures will be large in scale too, so what happens when a 10 hour journey leads to a dead end, the depletion of supplies, a converging ocean of safeguards? One of the things that makes Blame amenable to adaptation is that despair-inducing run-ending softlocks like these are thematically appropriate. If they couldn't happen, if that despair couldn't really be encountered, it wouldn't be a very good adaptation. Likewise, if the obstacles were not real, then the sense of transcendence in a good ending wouldn't be real either.