What this checks
This reads the rel=canonical link on the page, resolves it against the page URL so a relative value is judged correctly, and reports whether it points at this page or at a different one. It runs with the URL shape check, which flags the things that create duplicates in the first place: capital letters in the path, stacked query parameters, and paths long enough to be truncated when shared.