[Version 9.0] Feature support for covariant return types

[RexJaeschke]

This PR is my formal version of the corresponding MS spec.

The following issues are yet not handled by this PR and need to be addressed:

1. 12.8.12.4 Indexer access: The MS spec states, “A similar change should be made in §12.8.12.4 Indexer access.” However, no proposed edits were given.
2. 15.7.6 Virtual, sealed, override, and abstract accessors: The MS spec states, “We similarly permit override properties and indexers in interfaces as specified for classes in §15.7.6 Virtual, sealed, override, and abstract accessors.” However, no proposed edits were given.
3. 19 Interfaces: The MS spec states, “We will need a rule that an explicit interface implementation must declare a return type no less derived than the return type declared in any override in its base interfaces.” However, no proposed edits were given.
4. 19.6.5 interface mapping: The MS spec states, “This is technically a breaking change, … Due to this breaking change, we might consider not supporting covariant return types on implicit implementations.” We need to check if it’s supported.

[BillWagner]

@jskeet This is ready for a first read.

[jskeet]

The parts about searching through classes and interfaces have lost me, but I suspect that's just my lack of concentration rather than a fault.

[BillWagner]

• §19.6.2: Added "(if the member has a value return)" qualifier to the explicit interface implementation rule, preventing unsound ref return covariance.
• §19.6.5: Added "if A has a value return" qualifier to the method matching rule, same ref-safety fix.
• §12.6.2.2: Reworded interface parameter list text to say "most derived interface among the static type of the receiver and its direct and indirect base interfaces" — clarifying it refers to the interface hierarchy, not all implementors.
• §15.7.6: Added informative note about the property type accessibility constraint, paralleling the note already in §15.6.5 for methods.
• §19.6.5: Moved the covariant return type note from a standalone Note block into the "Notable implications" bullet list, matching jskeet's suggestion.
• §12.8.10.1 : Added interface type-picking text to the returns-by-ref bullet for completeness (identity conversion only, no covariance for ref returns).
  -§15.7.6: Updated sealed property text to say "derived class or interface" for consistency.

I still need to investigate bullet 3 in Rex's notes above.

For bullet 4: This breaking change wasn't implemented. Rather, covariant return types on implicit interface implementations isn't supported (including by derived interfaces).