Speed up convenience wrappers by allowing pre-parsed model

[maxconway]

Much of the time involved in model parsing can be avoided if a similar model is supplied.
Might even be possible to automatically cache models