From a recent email thread:

The first that comes to mind is that it doesn't tell you if the string was expected or not - i just says string found or not found, but not the setting of the "expected" flag. So it might say "string not found" but that was fine in the case where the string isn't expected. Until I realized that, I was interpreting the failure output incorrectly in many cases - since the test framework outputs all tests' output even when just one test in a class fails. This explains why I initially thought so many unrelated tests were failing!