> That said, the issue was kind of solved with, you know, hyperlinks and HTML.

How do you mean?

For each cited item (depending on a context; a case, a paper, etc.), one would provide either an in-text hyperlink or a hyperlink as a reference in a bibliography. Insofar as science goes, it would be easier for both authors and publishers if only DOIs were provided, as full and proper details could be retrieved automatically and hallucinated ones would be automatically caught. Additional identifiers (ISBNs, URLs, etc.) could be used in rare cases where DOIs are not available.