“Pollution” is a strong word that comes off as needlessly hostile. It seems
prudent and rational to get a broad sense of the landscape(and where it is
moving). The Wikidata data model is not trivially 1:1 with RDF/SPARQL and
there may be scope for hybrid solutions.

@Lydia_Pintscher mentioned a conversation with the Data Commons team at Google, they have this opensource codebase that's somewhat in this area: https://github.com/datacommonsorg/mixer

How about generating sitemap files during munging (rather than as part of
mediawiki or wikibase frontend)?

See also discussion on Twitter; various other search systems also don’t find all Q-pages.

yes, front page sounds good

Yes that minimal example is the minimum used for Google Dataset Search to have enough info to include a dataset in its collection.

From the quickest of looks (and totally unofficial w.r.t. my employer, of course): I don't see the string "entity" in https://www.wikidata.org/robots.txt so it may not be excluded there

from a quick look, I wonder whether http://schema.org/about would be a good fit. In most cases you're trying to characterize the topic of the page, aren't you?