The big fight about non-targeted vs. targeted input is a red herring. The fact is that it’s all CI. The only difference Tina and I are suggesting (Jan. of 2016) is, instead of working from words connected to a curriculum, with the focus on teaching the words, we instead work from an image, which results in a much richer focus on a whole unexpected range of linguistic possibilities that arise when we do that, which in my experience brings, to use Krashen’s term, far less “constraint on interest”. This is in keeping with this:
https://www.yahoo.com/news/m/11ab9279-35f2-3263-9ba0-82d2a7b497d5/ss_millennials-might-as-well-be.html
