I'm interested in seeing if anyone has experience to offer in extending an organization's existing data governance model to include open data release practices. For example, how have organizations utilized data governance roles (data owner, data steward, data custodian, etc) in the management and release of public data?
If you're talking data governance frameworks, I think you'll want to look at the Data Management Body of Knowledge (DMBOK) published by DAMA International. The functional point of view in DMBOK is pretty comprehensive. As Federal government guidelines on roles and responsibilities go, the best baseline framework is what the Office of Management and Budget (OMB) and agencies put together in response to the information quality act. OMB's guidelines are available here: http://www.whitehouse.gov/omb/fedreg_final_information_quality_guidelines
Every federal agency was required to develop information quality guidelines, but here's an example from the US Department of Transportation: http://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/subject_areas/statistical_policy_and_research/data_quality_guidelines/index.html
The guidelines generally ponder a data quality administrator and data quality officials matrixed across the organization. DOT adapted most of that goodness into its data dissemination policy, DOT Order 1351.34. Which you can find using a google search because I don't have enough reputation points to link to it. :-)
(Disclaimer: I am a contractor at the US Department of Transportation)
It's hypothetical but a great (and realistic) model is http://project-open-data.github.io/future-case-study/.
The Department of Health and Human Services (HHS) has experience adapting its data governance framework for open data. The effort has been led out of the HHS CTO's office.
Disclosure: I work for the Asst Sec of Planning & Evaluation (ASPE), an HHS agency. I work closely with the CTO's office. I have experience to offer though I was not originally part of the extension of the data governance framework.