The future of Dependent Haskell

This gets back to the post which (helped to) start all of this:

So where exactly is the “grand plan” for a “Dependent Haskell” ?