Good thinking, and would enable us to keep ideas like 'stickers' that we've kicked around internally as fun future features (that are way less hard in a non-responsive environment)

Sorry, I didn't write that very well. I think the consequences are pragmatic ones. In order to achieve more complicated visual designs you have to do a lot more work to make them considerate of all the different situations they might appear in. So pragmatically you either make less designs, or you limit the designs to simpler to scale paradigms (like block layouts) which require less pre-consideration.

As the foundation of this app is the idea that a user can just switch layout/visuals, it needs a lot of visual looks to switch between. Doing a lot of those, and having the responsiveness, and then iterating on app/data model features (which need updates to those designs) is quite a drain on speed.

All that said, since you asked for a single _actual_ example, I have one which I think is quite problematic (though could be addressed in some way, just not easily). We've kicked around the idea of letting users place stickers over the top of their design in order to customise them (a little like Instagram Stories), it's not obvious how to migrate a user input visual through a responsive set of designs without a lot of trouble (the most obvious solution being to guess and then let the user tweak at different responsive breaks).

Hopefully that convinces you that it wasn't just a throwaway 'nah we don't do responsive' :) I'm sure we could do better than where we currently are, and it is an iteration one release!