Models - usefulness and otherwise

Well, my last piece on Rhizomatic Learning.
In today's #change11 MOOC live session a series of questions allowed Dave Cormier to expand on the ideas and again it was an interesting hour.
At one point Stephen Downes started to sketch to explain traditional learning and rhizomatic learning (it was me typing in the labels - it may not have been exactly what Stephen meant).
All of a sudden it started to make sense, the idea that nothing new is created with traditional transmition of knowledge, that rhizomatic learning created something new, and then STOP! On the wrong track, Dave dragged us back underground to draw the rhizomatic approach (top right).

For me, a model has to be simple to follow and we should not need to read French philosophy to "get it". Stephen's alternative started to make sense, Dave's rhizomes clouded it for me. So, for me (and I stress that others may well find the concept useful) it goes no-where and as a model fails.
However, let me add something more concrete to the question raised about "assessment" or at least the badging or verification of such learning.
I can see that for teacher professional development, an open non-curriculum self-directed connected learning approach works (have I just described rhizomatic learning?). In my school, those at the cutting edge of technology for learning are learning in exactly this way. The ideas that permeate from them are proof that the learning is taking place. They are badged by their subsequent actions, and this is a perfectly acceptable verification.