Rename fhir-model code subtype nested classes from ValueSet to Value

[lmsurpre]

Is your feature request related to a problem? Please describe.

When we first designed the fhir-model classes for R4, we decided to model Code types with a required binding in the base spec as their own classes that extend Code and the bound valueset as a nested class called ValueSet.

However, as I've been working with these enums more and more, the name ValueSet has started to bug me.
Each enum value is like an instance of the Enum and so there is some dissonance when we define a member like this:
MyCode.ValueSet myVal;

Because myVal is actually a specific value for MyCode, not its ValueSet.

Describe the solution you'd like

1. Add a peer to the nested ValueSet classes and override functions as necessary to make the values equivalent to the existing ValueSet values.
2. Deprecate the old ValueSet classes and remove all references to these throughout the codebase.
3. After a couple releases, remove the old ValueSet classes; probably in combination with the introduce of support for Support for R4B #2068

Describe alternatives you've considered
live with the ValueSet name for these nested classes

Acceptance Criteria

Additional context

[kmbarton423]

No additional verification needs to be done with this one.