🏗️ Leverage `frictionless` and `pydantic` packages

[e-lo]

As a maintainer, I'd like to minimize the amount of code I'm uniquely maintaining and leverage active community-maintained packages.

Frictionless Framework Package: validates schemas and data to schemas
Pandera: can read, validate, and coerce pandas dataframes to frictionless schemas

GMNSpy was instantiated before the above would work, so there is likely a lot of duplication that can be resolved and the package slimmed down.

Should test, though, for speed.