The extract paragraphs does mostly what you want - can scan through one or all notebooks and only pull out/extract the text blocks that have the the tag all on one page (it's block oriented like Logseq linked references - not the whole page). It's not fully dynamic - but can do what you're saying manually on demand.