Some of his deductions are flawed: In the Nike example, he confuses implication with equivalence. Also, uncertainty can be modeled with probabilistic logics ("people in France speak French with probability .95"). But most if it seems still true. Ten years later I don't see any popular applications, RDF is not a very popular exchange format, and most of the available data is published by projects with the explicit goal of creating data for the semantic web (DBpedia).