Hi!
I have seen that SPARQL query service and it indeed is an interesting alternative. In terms of stability and update frequency how different is the SPARQL query service from the Wikidata Query API?
In terms of stability: it's beta, so while we try to keep it up and running smoothly, it is not out of the question that it can be taken down at any moment, either because we found a bug or because we need to update something, and the data model can change too. We do not expect substantial changes in data model anymore, and we try to keep it up and running (doesn't help that we are in the middle of large labs outage right now: https://wikitech.wikimedia.org/wiki/Incident_documentation/20150617-LabsNFSO... ) and synched continuously (i.e. no more than minutes behind wikidata edits), but as long as it's beta we can give no guarantees on anything. We're working hard to make it production-quality, but that will take a bit more time.
The differences between WDQ and WDQS/SPARQL is that SPARQL is a full-features language for querying triple-based (RDF) data sets, and allows very complex queries. It is also a standard in linked data world. You can use the translator (http://tools.wmflabs.org/wdq2sparql/w2s.php) - once the labs outage ends of course - to convert between WDQ syntax and SPARQL. Also check out other links on the WDQS beta page for short intros about how things are done with SPARQL and examples of which queries you can run.