At over 2500 lines long this is probably the largest use of the PDF/A extension schema definitions on the planet. Our own pdfaValidate schema has been updated too. It now includes some new properties ('default', 'subst', 'predefined' and 'count') to help us use this RDF schema to build our data-driven PDF/A XMP validator.
Feel free to use this data to build your own PDF/A XMP validator but remember to give back: if you have corrections or improvements, please share them with the PDF/A community. Better still, join the PDF/D Consortium.