The GRDDL test results are reported by developers in RDF using a the EARL vocabulary, i.e. Evaluation and Report Language. The tabulator is a Semantic Web browser.
If you load, e.g. GRDDL.py_2007-06-26.rdf into the live tabulator, you can browse around the test results:
EARL test results are associated with assertions:
Note the tabulator is loading relevant schemas and such as you browse:
Or load raptor-earl.rdf and get a table of test results:
Load those two plus jena-earl.rdf and get a table of results from all three:
Then perhaps you're curious about some particular rule, so you can explore it. Note that the GRDDL tests are indexed by rule/feature, so you can explore those too: