I’ve found them useful, both the more accurate Japanese concept of the things and the WK ones. Mostly because they help me chunk the characters better. This works both in terms of like…memorizing them and in recalling them while writing. So then it’s not just like…3000 completely unique blocks. It’s more like…3000 blocks comprised of set bits and bobs in various combinations. That’s a much more doable task for me, since then it’s not just Unique Individual Characters it’s…small blocks that are built into bigger blocks in greater number.
I think the most useful thing for me though beyond that with like traditional set of them and not the WK ones is being able to look up kanji by radical. So I think it’s very worthwhile to at least learn the most used of the not-WK ones so that you can utilize the radical indexes in kanji dictionaries and such to find stuff quicker. If I come across an unfamiliar character I can go to like jisho or look at my kanji dictionary and look up the radical and find the character. Convenient! That can help with figuring out words you don’t know the readings off too. Since then you can look up the individual characters and like…figure out the readings? Or you can like brute force it and find the individual kanji and then smash them together into a dictionary search and bam. Word is solved.
In both cases, the key thing is creating a more manageable to search index.