Possible misbehaviour of some tests

I'm preparing the reference update needed by gaudi/Gaudi!1426 (merged) and I noticed that the change in the standard output is not marked as a failure in x86_64_v4+vecwid256-centos7-gcc11-opt.

Looking at the tests that are supposed to fail according to the results for other platforms I noticed a strange thing in the validator:

# Check there are no NaN values in the ntuple
from DaVinciTests.QMTest.check_helpers import has_nan
assert not has_nan("DV_example_configFuntuple_ntp.root", "Tuple_B0Dspi/DecayTree")

I didnt't check the code of the test wrapper in Gaudi, but it may be that a failure of the assertion marks the test as passed (by accident, of course).

I believe that in the current form all those checks are void.