From FRBR 3.3, Aggregate and Component Works--"... from a logical perspective the entity work, for example, may represent an aggregate of individual works brought together by an editor or compiler in the form of an anthology, a set of individual monographs brought together by a publisher to form a series, or a collection of private papers organized by an archive as a single fond."
My view is that most edited editions of literary works are aggregate works per FRBR because the combine different kinds of work-level content, but that we don't see a need to describe the aggregate as a separate work. A description of the manifestation with access points which reference the primary work contained in the aggregate and that work's expression-level contributors (editor, translator) suffices. But as Bob Maxwell pointed out, art and an editor's introduction in such an edition would be regarded as separate works, not part of the primary work's expression. If the artist is named in an AAP, that's properly part of the description of the art as work, or the description of a component implying the presence of the aggregate work, though the aggregate work per se is otherwise not described in most cases.
Just because we don't pay much attention to the aggregate work in most cases doesn't mean it isn't there. In the case of series, it may be useful to pay more attention to it.
The question of expressions of the series-as-aggregate-work gets back to my initial question. If the series AAP represents the aggregate work, and a publisher releases an alternate language edition of the series (same art, same introduction, but in Spanish) as may have happened with HarperCollins Spanish edition of Narnia, then that expression of the series work should probably take the AAP "Lewis, ... $t Chronicles of Narnia (HarperCollins (Firm)). $l Spanish" rather than "Lewis, ... $t Chronicles of Narnia. $l Spanish. $s HarperCollins (Firm))". The latter would be the correct AAP, distinguishing a new series aggregate work (not expression), only if the component works (art, introduction, etc.) which accompany the text are new, and not carried over from the English edition of the HarperCollins series.
At least, that's how I'm inclining at the moment as I ponder what to do with my Narnia in Danish. I second the list of concerns that Mary Jane Cuneo adds. My goal is to formulate principled FRBR/RDA basis for continuing the practical solutions to these questions that we've used in the past when creating series authorities; but other interpretations are welcome.