Review test results prior to "official" report publication

[steve-bate]

Do a final pre-publication review for test results, looking for:

• Incorrect status (due to framework bugs or behavior, for example)
• Misleading results (result "shadowing", etc.)

This should be done shortly before publication to minimize regressions from code changes after the review.

[jernst]

I did this recently but we should do it again for release 0.2