Sadly, probably not.
Despite a few bits of feature-creep (like TOTP support), I've been trying to keep passlib primarily focused on password hashing. NaCl is more of a general-purpose crypto api, which is a larger surface area than I've got the time to support.
Also, a secondary goal I have is to keep passlib written in pure-python; conditionally taking advantage of any third-party C extensions that may usefully accelerate parts of passlib (e.g. M2Crypto, py-bcrypt). NaCl doesn't appear to currently offer any password hashing algorithms, or any other bits that would be useful as yet; but if it does, and another project provides a Python API, that's probably when / how any NaCl support will be integrated into Passlib.