Blanket rules about the acceptability of sources set off WP:IAR constantly. Blogs are a case in point. TNH's blog entries about herself are clearly primary sources and would tend to be preferred and in some cases necessary.
Blanket rules are a problem when one has a lot of amateurs writing, When you have to tell people which types of sources are acceptable, it's almost a foregone conclusion that they are going to fail to deal with the evaluation of the sources they have fairly frequently, because they are likely to rely on authority too much. For example, I've been doing lighthouse articles of late. Basically I've been using four sources: the USCG website, two lighthouse society-type websites, and a sort of coffee-table book of photography. I'm also checking them against Google Earth and NOAA charts as to the locations given. Which source is the best? Well, all of them have been caught in some sort of conspicuous error. As it turns out, the coffee-table book is generally the best-- and as it turns out, everyone else uses it as a reference too.
Yes blogs are generally not a good source, but it's extremely dependent on context.