Ok, this was simultaneously a spectacular failure and a marvellous success.
I went for a walk and I came to some conclusions about how game design works, and came up with some interesting questions.
Firstly: I enjoy playing this game. Why would the dev enjoy it, while some players suffer ridiculously?
Why is the dev's experience different from the players' experience?
Secondly: most people hate Getting Over It, but I loved it. Why the heck would that be?
Why should players' experiences differ?
There's two concepts I think I've come up with here, or at least that I've come to a better understanding of.
Firstly, on goal structures. The dev has a different goal structure than the player: the player is looking to have fun (whatever 'fun' may be, I am not yet sure), while the dev is looking to /find/ the fun. So long as the dev finds fun /somewhere/, they will overlook the difficulty it took to get there. If the player isn't immediately presented with the fun, they'll be more likely to give up.
Regarding Getting Over It: streamers have a goal structure that allows them to get through the game, because they're rewarded for being frustrated (by their audience). I think I personally went through all the work of playing Getting Over It because there was a secret at the end, a relatively well-kept one. Because I had the goal structure of 'there is a reward at the end, and it could be anything,' I overlooked the difficulties inbetween. But the /internal/ reward I gained from Getting Over It - i.e. when I got a dopamine kick - came not when I saw the reward itself, but when I made progress /towards/ that goal. Now when I play other frustrating games, I see in it the same potential for reward that Getting Over It gave me.
Here's the second concept I better understand now: actions and meta-actions. The lowest-level actions of any video game are those of pushing buttons, and higher-level meta-actions are like "move the character to the end of the level" or "fling the character from this node to the next one." As the dev I was able to get past the lowest-level actions, and found the /fun/ in the higher-level actions. I became blind to the difficulty of the low-level actions.
Here, I think, is how this game failed: some players had to face off against higher-level meta-actions when they hadn't yet mastered the lower-level actions. Specifically, >>36746388
clearly needed to be introduced to the individual actions of 'set the neck at this specific angle before jumping' in a safer context, before being asked to combine that with other mid-air moves.
Also, I think the 'skip level' button was a bad idea: just as not knowing what was at the end of Getting Over It gave me the motivation to get through it, knowing what's at the end of the game can take away all your motivation to complete it. There's still a 'secret' """true ending""" that people can reach, by the way. It involves dr. banana, if you can find him.