Various social network silos (e.g. Flattr, Google) has or has had verification of third party content on their sites.
Flattr uses internally mapped URL:s to known proprietary identities and then used proprietary OAuth API:s of other services, like Twitter and GitHub, to let user verify multiple identities per platform.
Google's "Authorship in web-search" feature used rel-author and similar to establish authorship of a web page and then tried connecting that to a Google+ profile, either directly or indirectly by following rel-me links to and from the authors profile and the Google+ profile. Google then presented a profile of that user in connection to the page in their search result.
Twitter offers a subset of users the option of a blue "verified" checkmark icon on their name/profile, intended to "establish authenticity of identities of key individuals and brands on Twitter."
Authorship + Rel-Me
First determine the authorship of a web page. If the author profile is on the same site (same host name) as the web page, and thus can be expected to not have been spoofed, then check for a bi-directional rel-me link between that author profile and the identity profile to verify it as.
To enable verification of more complex rel-me identity links one would need to compile an entire identity graph, the identity's own subset of the full social graph, and traverse the identity graph from the author profile, to the target and then back and ensure that the full chain of links exists and are valid. This would require something like identengine or RelSpider .