Log requested document IDs that are not found in the API backend

[bill-baumgartner]

In an effort to ensure we are maintaining a stable and up-to-date collection of document metadata, it would be helpful to log document identifiers that are requested by users but that are not present in the API backend. Is this feasible?