Is it useful to target property names, types, and tags with new methods in this library?

E.g.

• A tag targeted by AddTagProvider which is an alias for AddProvider
• a type by AddTypeProvider (reflect-based)
• a field by AddFieldProvider matching field name?
• leaves room for other types of strategies later

It would probably require providing different "targeting" strategies for determining faking values. I understand there'd have to be precedence for applicability too. I think each of these could be valuable, though I don't know if its considered applicable here.

I can provide some implementation/PR if this is within scope. I'm going to try things out tomorrow and come back with some more details.