In principle yes, I agree that former employees' sentiments have an obvious bias, but if they all trend in the same direction - people who worked in different times and functions and didn't know each other while on the job - that points to a likely underlying truth.