I will try and contact you by mid-May to discuss this.
One of our members took a look at the latest Canonical XML 2.0
draft. Although it appears to represent a leap from 1.0, it does
not seem to address our desire to bypass altogether the textural
XML representation in getting to the final hash value for signature.
We, therefore, are considering that we may want to work on
native EXI-canonicalization in the next WG term.
I would like to know if there has been any discussion or interest
had by XML Security WG based on our write-up  so far. I think
that if there are enough interest, the two WGs would better be
working on it together.