To what extent should we describe NTRU

[sfluhrer]

One issue is that, without NIST approval, there is not really definitive definition of NTRU. The closest would be the NIST Round 3 submission, however the description there is targeted towards cryptographers/mathematicians, and I believe that a standard security engineer would not be able to understand it. Should we try to write up an equivalent definition (targeted towards engineers without extensive math background)?
Should we include reference code (either in addition to or instead of the above description)? Or, should we point people to the reference code in the Round 3 submission?