Maybe. I suspect there might be some value hidden in the aggregate data on Glassdoor (but that really is just as susceptible to the same type of manipulation). The number 5 may also be significant or insignificant depending entirely upon the size of the organisation.

But really, if your reasoning was robust, it would work in both directions. I don’t think you’d agree with the statement “if five people say that their work environment is great, there may be something to it”. The system can be (and is) gamed in both directions, and there are people who have incentives to do so in either direction, for both “justified” and “unjustified” reasons. Any reason you come up with to be skeptical of a good review also applies to a bad one.