If one were to subscribe to such a distinction between algorithmic ranking and algorithmic suggestions I would liken it with a broad paintbrush to:

Ranking: A group of people share a ouija board, and together make selections.

Suggestion: A singular entity clips together media to create a new narrative, akin to a ransom note.

If the sum of the collection of content more than its parts, if is different in strength not kind, or self reenforcing, it's really hard to distinguish where the algorithm ends and the voters begin.