WaniKani effectively just pulls Japanese words that provide reinforcement for the kanji out of a dictionary. Just because a word is in a dictionary doesn’t mean it is used commonly (or at all) or doesn’t have a contextual limitation. Choose a day to tell everyone “I’m feeling quite gay today” and see how many don’t, at first utterance, question your sexuality.
There are a lot of cases in which some words have a nuance or specialized definition for them that wouldn’t be easily recognizable. The first time I used 語る to refer to “telling a story” because my students kept writing “talking a story,” my JTE informed me that 語る is rarely used because it is prideful in nature, so they use 話す instead.
A great example of this would be the now-removed 先年. I had to explain to natives what I meant by writing the kanji because when I said “せんねん,” the only thing that came to mind was 千年. Many of the words aren’t MEANT to be used normally. Since there was plenty of vocabulary to cover the readings of both 先 and 年, having a word so unusual now was pointless. On the other hand, your friends understood your usage of words, meaning they are both useful for reinforcement and recognizable (albeit either situational or strange colloquially) in conversation, so there’s little harm in learning them.
Personally, I love those situations where natives clue me in on those types of things, because then I can research why one is better (or more correct in that situation) than the other. I think one the biggest broad examples of this is all of the Chinese/Japanese synonymous verb pairs. It’s both frustrating and fun to figure out which to use (experience has indicated typically to use the Japanese unless their is a specific nuance or grammatical difference like 理解する vs 分かる), especially when they tell you to use the other but then can’t really figure out how to tell you why it’s better.