What if, in effect, you have two “shifted” modes, resulting in the possibility of 4 different states? Is this worth avoiding in favour of simply 3 states (one per button, plus “idle”) which can interrupt each other?

I’m going to intuit “yes” because of some experiments in videogame controls I’ve been doing over the years. In my experience, it’s best for the “youngest” quasimode to interrupt any other mode (a good example would be “ironsighting” vs. “sprinting”) and for negative edges to revert back to the mode which was interrupted, if the key for that mode is still held.

i.e. it’s generally a bad move to have quasimodes which lock out other quasimodes (there are some examples where it’s useful to have this, but they don’t spring fervently to mind) and even worse to have two shifts lift you into a compound quasi.

I am not being very good at giving examples or evidence for why I feel this way. I was hoping you might be able to tell me whether or not I am mistaken, and why?