Forecasting open data demand is quite challenging. A new public dataset means more downloads, a new app results in way more requests.
It would be interesting to discuss the throughput of open data portals, in terms of datasets downloads or API calls.
Has anyone done a research in this matter?