Obviously a group of people would not be satisfied with any given company's service - especially in the case of Afrihost who by their own admission had a difficult year thus far. This dissatisfaction is to be expected, and it should be no different.
I can only assume, however, that many more people were indeed satisfied with the service they received. In the end, a survey is about averages - there is no point in absolutizing one's own negative experience with a service or company.
With regards to the suspicions about the integrity of the survey results: I highly doubt anyone care enough to actually spend time on skewing the results in favour of a particular company.