Another point in agreement with joethdo: strategists who predict the end of DVRs, while they may be right long-term, rarely mention the near- and mid-term issue of speed and ease of use.

Even as more content becomes available on the Web for streaming, the user experience compared with watching a downloaded file isn't yet the same. Netflix is great until the user tries anything other than play. Pause, fast forward, jump back, etc. are not nearly as responsive as when played from a DVR. There can be a several seconds delay, or it can be nearly impossible to have the same degree of control as when played from a DVR. This is an advantage of any play-from-local-recording service, whether a recording from live tv or a download from Amazon.

So, in the progression away from DVRs, it's not just a matter of making all that content available. It also an issue of server and network speeds, etc.

Of course, a user might have to have experienced the speed and responsiveness of TiVo play from recording or download in order to appreciate what's missing in the Netflix and other similar streaming experiences. Maybe that's why these authors rarely seem to mention this issue -- they don't know what they're missing.