What is the standard error on these measurements? And what is the variance, that is does the mean satisfaction represent the majority or are there, say, two populations, one very optimistic and the other less so.

Perhaps that could explain the outlook seen on HN, etc.