Colored result interpretation

[Bender250]

When some of the tests in battery fails with extreme p-value (1e-300), signalise it for the whole battery with (for example) blue background with meaning "You might want to rerun this test for different seed and find out, if the same test fails again".

Example: http://147.251.253.249/ViewResults/Test/175856/

This is repeatedly failing test 13. But it is a single failing test, so the battery interpretation is green "OK". But we can manually reject the randomness hypothesis.

[Bender250]

After discussion with @sysox, we should highlight that result with red background, as 1e-300 is extreme p-value. Maybe some interval like 10^-5 to 10^-7 can be other color?