Ad-hoc querying affects all distributed databases, which includes dynomite, hbase, and hypertable as well as voldemort and cassandra. Also all key-oriented databases, which include Tokyo Tyrant and redis. And some others like CouchDB (where computing a new view via map-reduce is basically equivalent to doing a query via hadoop against another system, so I don't consider that real ad-hoc query support).
Really the list of databases that do provide ad-hoc queries is much shorter. Neo4j (graph db), mongodb (single-master "document-oriented" db), the xml databases, probably some other minor projects.
TFA also mentions data export as a "dark side" but for projects that lack this it's really just a "nobody's needed it enough to write it yet" thing. So that's less interesting IMO than talking about ad hoc queries which is usually more of an architectural limitation.