It’s interesting to hear them, and personally I want to go in the opposite direction from both of these suggestions.
Take Kerbal Space Program as an example, it will let you put 100 space station components on a tiny booster and take it to the launch pad. However when you launch it will explode. This is a great moment of learning, you have to think, and in the end you figure out that a rocket needs a Thrust to Weight ratio >1 in order to launch. That is a great moment because you have learned a universal law of rocketry and it feels good to discover it.
Same with Thrive. I want people to make crazy cells where it’s one hex of cytoplasm and a nucleus and have the species numbers plummet and maybe die out. This is the kind of failure which will teach you about osmoregulation and square cubed law, which are great things to learn.
Likewise I want people to try and use chloroplasts in dark caves. I want them to think about how light levels (and maybe even spectrum) impacts chloroplasts.
In some sense I want to make it easier to fail so the game is about learning.