There's a somewhat lively debate going on in the
COW Final Cut Pro forum about vDSLRs (still cameras like the Canon 7D or Panasonic GH1 that have HD video recording ability) and their viability as HD video cameras vs more "traditional" HD prosumer level video cameras. A few reasons why the vDSLRs (or HD-DSLRs depending on which acronym you prefer) are a hot topic is because of the low price, interchangeable lenses, and shallow (film-like) depth of field that has been sought after by so many people using sub-$10k cameras. Some downsides to vDSLRs are non-standard frame rates (7D withstanding), need for dual system sound, poor ergonomics for video shooting, conversion from H.264 etc.,.
In this current generation these cameras are still cameras first and foremost with the HD video recording ability kinda tacked on. The original intent was for photojournalists who, by changes in their industry, are getting pushed into taking video for the web but DV Rebels have pounced and there are numerous examples of good looking video from these cameras (
Philip Bloom for example). Given how much people are typically gushing over these images an interesting note is that according to an article by Barry Green on vDSLRs and aliasing these cameras, when recording video, only resolve detail at a level somewhere between SD and 720p video. When I read the article I found that truly surprising because of how much people were lauding the quality of the vDSLR footage. A common retort is, "I don't shoot test charts for a living" and if the footage looks good in the end does it matter what the test chart says? Yes and no, IMO. The test chart gives you an objective, standardized look at how the camera performs which can be helpful in choosing the right camera for the job, but it shouldn't be the end-all, be-all of your decision making process.
So, in the end, horses for course, IMO. A vDSLR might be perfect for some and horrible for others and that's okay. The sub-$10k market is so packed with cameras that have different strengths and weaknesses that with a bit of research almost anyone shopping in that price range should be able to find a camera to suit their specific needs. I think that video and still cameras will continue to converge and it will interesting to see what Red's Scarlet does to this market place when it lands sometime next year. Scarlet will be unique in the playing field in that it was designed from the ground up to be a still *and* motion video camera (as opposed to a DSLR that can record movies or a camcorder that produces good stills).