follow-up to omitting types in augmentations

[jakemac53]

Follow up to some comments on #3940 after it was merged

[lrhn]

This is tricky stuff. I have ideas. I should try to collect them. 😄

[lrhn]

"was already present".

This is an error that should probably be reported during augmentation application.

When applying an augmenting class declaration to a class definition (as defined by a base
declaration or a number of prior augmentations applied to a base declaration):

It's a compile-time error if the augmenting class declaration contains an extends clause and

• the declaration is a mixin class declaration, or
• the augmented class declaration already has a declared superclass.

If the augmented class definition has a declared superclass, the resulting class
definition has that superclass.
Otherwise, if the augmenting class definition has an extends clause with a type term S,
the the resulting class definition has S as declared superclass.
Otherwise the resulting class definition has no declared superclass.

[jakemac53]

I did reword this a bit, but I think it is all covered by other rules in a general fashion. Above, we say:

A class, enum, extension type, or mixin augmentation may specify `extends`,
`implements` and `with` clauses (when generally supported). The types
in these clauses are appended to the original declarations’ clauses of the same
kind, and if that clause did not exist previously, then it is added with
the new types. All regular rules apply after this appending process, so you
cannot have multiple `extends` on a class, or an `on` clause on an enum, etc.

[lrhn]

(Don't know where we define it, but it's probably correct that a function signature differs from a function type in that it can contain covariant. It's a little iffy whether we can talk about non-interface-members as having a function signature, though.)