Let’s say for example I typed ‘left’ for meaning of 右. In this case 右 will not go out of review queue(as it means ‘right’), but it also clear that I not only don’t recognize ‘右’, but don’t remember that left is actually written as ‘左’, so it seems reasonable to review ‘left’ as well asap.
So I suggest when a word is failed, wanikani should search all learned vocabulary and if there are words with entered meaning, add them to review queue (all or one at random).
A similar idea was once proposed where (using your example) all other vocab containing the kanji 右 would also be marked wrong. However, think about it this way. What if you wrote “stone”? Would you still have 左 put up for review again*? What if you typed “loft”? It’s not technically “left”, but it’s also not unambiguously clear if it was a typo or a completely different mistake.
Basically, the way that thread went is that there are too many factors into why an answer is incorrect to penalize the user for other items. If you think about it the other way, if you got 右 correct, that would imply that you also know that 左 is left, which isn’t necessarily true. And this is a very neat-and-tidy example - related kanji are usually really messy!
Sounds like something that could be integrated with the Confusion Guesser script. Manually go over and look for the one you mixed it up with and mark it to drop down an SRS level or two, though I’m not sure if WK allows this functionality. I’ve never seen a script do something like that.
Also I noticed this is your first post. I hope we aren’t coming off as too aggressive, and we welcome you to the community! We all have opinions on how WaniKani can and should be done haha. Best of luck