Are the small examples given in the DCAT spec supposed to be building up a more complete picture of a bigger example?

If yes, then I'm puzzled by the story told for dataset-001. In example 6 its primary type says it is a dataset of manuscripts. But its content type in example 7 is "standard office documents", which hints these were not created purely by hand. And its temporal coverage elsewhere says it covers 2006. It's unlikely manuscripts would be created as a main source of data at or after that date. And it has a distribution as csv files, later.

I reckon that it is meant to be an imaginary dataset. Still we could aim at presenting slightly more consistent chimeras :-)

This is probably for future work anyway.