THIS is for you, then...

And the reason is pretty simple: the approach to use is "different"... (and don't discuss that sentence here, please ;) )

I find it a pity that boundaries with 2 consecutive corners (++) are not tested; I see no reason why not, except that it would break some already submitted solutions.

Strongly agree with this one. as it make the tryCode to be a simple function without the need of caching the previous guesses.

why noone is xor'ing? kinda dissapointing :(

Yes, that is guaranteed and the latter is the minimum shape.

I consider it recursive if the same source code is being called with new data. this is a parameter, too. In fact, from an OO-perspective, I'd go even further and say it's recursion just to message-pass the same sort of message you received (except maybe in the case of classes that just decorate enough to log stuff).

I don't think there really is any recursion - the functions being called are owned by different object instances.

Is there a way to enforce a linear time solution, short of providing an input so big that a non-linear solution would cause a timeout? I would be happy to write a new one that requires linear time! (That's what I found so fascinating about this problem; haha, I'm even still trying to understand the math behind finding two missing variables in linear time...)

Awesome idea! I'll write up a Part 2 this week. I'll throw in a thanks to you in the description if that's cool with you :D

No worries, as long as it works :)
I think that's a great idea, but I wouldn't want to break all the solutions. Would a Valid Parentheses Part 2 be better?