At the moment you can't really see why a test fails unless you download the log.

I can see two ways to make this better.

1. Run the failing tests again at the end of the testsuite run.
2. Pretty print the junit.xml file using xq or something similar.