Idea about prospective memory

This is a kind of thoughts inspired by Nailler's posts on prospective memory and the fact that I find MILD "cluttery".

What would happen if you based your lucid dreaming activities on prospective memory instead of dreaming?

1) So when you wake from a dream, instead of looking at dreamsigns (or lack of dreamsigns) and general dream content, you specifically look for and log moments in the dream where you used prospective memory. What would the definition of PM in this instance be? Not sure, but maybe you could look at moments where you remembered to do something (like PM in real life), or perhaps also where you realised you intended to remember to do something in the future.

2) Going back to sleep, you assert your intention to use prospective memory in your next dream. And, perhaps, as consequence, in the next dream, you remember your intention to use prospective memory which is more or less the same as realising you are dreaming.

The transition from looking back at PM use in the past dream to remembering to use it in your next dream seems smoother and less clunky than the activities carried out in MILD (Recalling the dream. spotting dreamsigns, imagining dreamsigns showing up in the next dream, saying, "I'll realise I'm dreaming" six thousand times before you go back to sleep etc.) The activities for MILD seem like some kind of Occam's Razor, where there is a great number of unrelated tasks that can't segue smoothly from one to the next.