Both Kevin Burton and Matthew Hurst are a bit suspicious of David Sifry's recent state of the blogosphere report. Count me in as well. Burton delves into the number of posts per blog, among other things, and through a couple of calculations argues that the number of "active" blogs is a few millions and only growing linearly. The comments on his post are also worth reading.
Of course, this hinges on what do you call an active blog, which is where Hurst comes in. The blogosphere, if you conflate that with Technorati's index, seems to include a lot of dead stuff. Hurst also points out, more accurately, that Technorati is just one sample of the blogosphere. The hard part is how to extrapolate from that to a comprehensive picture of blogging. Studious readers will note however, that Hurst works for a Technorati competitor, just to be clear, and I've briefly met Hurst in a professional setting. I definitely lean towards his scientific approach to observing and analyzing the weblog ecology.
From my point of view, Sifry's report really calls out for some segmentation based upon blog activity. I'm not sure how productive "blogosphere is doubling" announcements are any more. They're akin to Google and Yahoo!'s index size announcements. Great! But how come I can't find that page I was searching for.