When data is bad or can be shown to be bad then a general distrust in ALL data will result.

In situations where the data is good but is telling an unpalatable story then the assumption will be that the data is bad and can therefore be ignored. Data is deemed to be guilty until proven innocent and no-one tries too hard to prove it innocent.

I have been in situations where man weeks have been lost doing a forensic analysis of a system to determine why the system was broken, how, who was responsible and what it would cost to put it right. Well it wasn't broken, just presenting an unpalatable truth.