It seems to me that the vast majority of newer software comes to market before it has been robustly tested and so the feedback agents permit collection of data that can be used to improve the quality of the software.  Perhaps I am overly pessimistic, but I highly doubt whether the designers or maintainers of the current Microsoft systems ever look at the data collected.  I believe that they are rushing to market with the next set of incompletely tested software.

As a database administrator I have seen collection systems for software problems for years and the volume of data that is produced is truly spectacular, but unless it is carefully organized cannot be aggregated to be useful for reference or investigation.  It seems that the time spent by the user entering the feedback might be better utilized restarting their machine, but perhaps may feel utility in communicating with Microsoft about the problems that they have experienced. 

In a perfect world there would be feedback loops that notify you that the issue you have commented about; is a recognized issue, has a name that can be used to refer to it in the future, and finally that if you like you can be notified when a fix is available to resolve the issue encountered.  The fact that the feedback loop is incomplete challenges the notion that we are working on systems of collective knowledge, but at best only with the appearance of cooperative and collective knowledge.