this is an excellent point. There are two points to consider here. GODI evaluates the performance of governments to provide data openly. Thereby we take into account that government sometimes might not be the immediate producer of the data, for instance when a public sector activities are transferred to public-private partnerships. Our main criterion is that governments should be able to improve their performance through any form of action - be it by publishing data or by mandating their publication. Thus in order to qualify as open government data, you should be able to find an indication that national (or sub-national) government at least mandated the publication of the data.
In the Measurement and Accountability working group of the International Open Data Charter we are trying to metrify Charter Principles such as “Open Data By Default”:
The term government data could also apply to data created for governments by external organizations, and data of significant benefit to the public that is held by external organizations and related to government programs and services (e.g. data on extractives entities, data on transportation infrastructure, etc.).
For instance we want to better understand the governance behind open data policies and implementation and to better understand single elements mentioned in the quote above. We acknowledge that the ecosystem of open government data may be more complex, and understanding how responsibilities for data production and publication are distributed might help us to refine our assessment in the future. At the moment, please go with the instructions I mentioned above.
Your case is very interesting and I would love to learn more about it. Maybe you can add the links to the data sets in this thread?
Don’t hesitate to send any other questions you might have!