By this milestone, jzon shall have a viable set of features for a version 1.

jzon won't be "Just Another CL JSON Library"
I want to be able to, without caveats, suggest to anyone using any other CL JSON library to instead use jzon.
Until that point is reached, v1 will have to wait.

The only caveat I'm willing entertain conceding on is possibly performance for very high volumes of data, where a 1x speedup is "Huge".
However, I suspect that performance within 10% of jonathan & jsown is possible without sacrificing jzon's other goals, namely safety, correctness, and convenience.