Building on the OpenSource CKAN project is a good thing. Similarly, you should thrive to help people build on OpenData.
What really concerns me is the actual data that's made available, more than the technology used to serve it.
For the time being, it looks more like electronic document management than OpenData to me. The philosophy behind OpenData is not just "let's put office documents online", it should be more "let's see what they'll do with our data". Otherwise, where's the innovation? (except for the ideology)
The three stated objectives of France's open data policy are: 1. to make government more transparent and accountable, 2. to improve public policies and public service delivery, and 3. to enable entrepreneurs and civic innovators to develop create new services based on this data.
That's why we've brought together more than 30 key players in the French innovation ecosystem (VCs, angels, research institutions but also media, larger French corporations, Google, Microsoft etc.) and organized a series data-driven startup contests called DataConnexions.
You can check the best projects from the first 3 editions (from early 2012 onwards): http://www.etalab.gouv.fr/recherche/?query=dataconnexions
My personal favorite is an app developed jointly with the Greater Paris transport authority that uses context-based modeling to predict train occupancy levels at each time of the day, in real-time, taking into account any perturbation on the network and live feedback from users: http://tranquilien.com
