feasibility of dropping Turtle=MUST from WebID ED 2014

Been having some interesting chats with Kingsley, and one route that he has
suggested to get to our goal of a modern webid set of specs would be to
drop Turtle=MUST

I personally see this as something that would be hard to gain consensus on
because:

1. The spec is already in use
2. Solid is at least loosely based on this work
3. Dropping MUST means that there is no guaranteed interop, potentially
degrading the UX in some systems
4. Timbl might not be in favour of such a change

I could be completely wrong on any or all of these things, but I thought
I'd ask the group, so that we spend our time on the things with best chance
of proceeding

If we can rule this path out then I think we have a relatively easy task of
defining a webid, writing examples and tests, then have multiple profiles
based on that shared understanding, of which WebID ED 2015, can be one

In short, I think we're 90% of the way home after defining the key webid
terms in a stable place

Just wondering what folks think on this path, if removing turtle=MUST it's
a non-starter, it would be good to establish early.  And that gives us
better clarity to get the modernized definitions and spec over the line,
modulo a few compromises.