I think the reason for shifts between Poser and DAZ|Studio has a lot to with increased familiarity with 3D applications.  When I started using Poser, there seemed to be a large learning curve in using the software's applications.  Now, people are coming into the community and picking up skills rather quickly.  I think a lot of this is due to the fact that the tools are much more advanced and the content is much more developed.  The relationship between the development of the software and the development and complexity of the supporting content is becoming more and more interlinked.

Over the past year, I've become a little disenfranchised with Poser.  Having Poser passed from one company to another makes future development rather questionable.  Poser is certainly superior to the others in terms of materials, but in terms of computer resources, I think there is significant room for improvement.

Further, I question the development of the "default" Poser figures.  With each new version of Poser, the figures have been renamed.  This practice doesn't necessarily establish a successful brand and maintain interest for the figures.  The G2 figures should have been combined into only 1 male and 1 female with more morphing options and better attention to the mesh.  The supporting content provided by Sixus1 and RuntimeDNA has been better than the actual figures.  In truth, I think Sixus1 and RuntimeDNA have given Poser more of a pulse that it's owner company.

Personally, I think a lot of DAZ's success is due to the fact they've established a stable brand.  The company has retained the same name.  Further, each subsequent generation of Michael and Victoria have become increasingly more developed.  Offering DAZ|Studio for free has been a great means of further advancing the development of the program.  I hope that eventually nosed-based materials are added - at that point, my usage of Poser might lessen even more.