It's true that some sites provide some sort of status blog or dashboard -- more, I now see, than I'd realized. However, most of them provide very little information -- often no more than "service is currently [up | down]". A few go so far as to publish some sort of graphs (e.g. Twitter, Google App Engine), which is refreshing, but even here the amount of information provided is fairly limited.
Third-party monitoring (e.g. http://api-status.com/) is another interesting angle. There is more that could be done here, but ultimately the best information can only come from the service provider themselves. (As a hobby, I've been doing some monitoring of AWS and Google App Engine for the last 18 months, but have not done much to make the results accessible. I've occasionally blogged about it, e.g. http://amistrongeryet.blogspot.com/2010/04/three-latency-ano...)
Another approach would be crowdsourcing -- users of a given service could band together to aggregate performance measurements. I discuss this idea in http://amistrongeryet.blogspot.com/2011/07/cloudsat.html.