Also, the first test did not reveal any kind of flaw in design or construction of the stage. It aborted early because of hydraulic pressure fluctuations that were within spec -- but exceeded extremely conservative limits in place for that test to protect the stage. The main thing that differed in the second test was a software tweak to relax those limits.

I personally think continuing the program is throwing good money after bad, given the availability of alternatives and the limited missions left for SLS (I'm not sure they're planning to use it for anything other than Orion at this point) -- but test failures are expected in rocketry (SpaceX has plenty in public), and it makes no sense at all to beat NASA up for this one.