So here I am learning vocab. After running through Core2k I feel I know the sentences well but that is only because I have seen them so often. A vocab word by itself I will not be able to get. Therefore, I made the decision to do vocab by itself. It is working.

However, to help with learning, I am doing a sort-of i+1 idea.
I took the Kore set of 6000 sentences/vocab and suspended all cards.
I then took vocab that I knew 100%. Easy stuff that I see and know completely.

子　女　原　電話　力

I unsuspend those cards.
Then, when I feel like new cards, I type in the kanji of one I am familiar with and unsuspend the vocab that contains that kanji and others I know and some I dont.

I want to add new cards with 子

電子　女子　原子力　女の子 子供

In this example, I knew all the kanji but 供, so that would be the new card to search with after I am done with the previous set.

As you can see, this repeats a lot of kanji that I know, reinforcing the readings of what I do know while adding parts that I dont. I continue doing this until theres no more kanji in the search or the ones that I see are too difficult.

Its a very simple concept and it seems to be going well.
Having said that, I was curious if anyone has tried this, or if there are any decks out there that use this principle? I dont believe I have seen a sort which does this - the sorts seem to focus on frequency rather than proximity.