I haven't had the privilege of seeing the report as yet, and naturally different media are reported as performing differently, but the overall conclusion makes sense to me and is consistent with the inevitable consequence of the growth of media that offer more control to the consumer.
VSS assigns the change to the greater efficiencies offered by digital media, greater bandwidth etc., but I think there is another factor likely at play here. After all, the more control consumers have over what they see and when they see it, the more purposeful their media consumption is likely to be.
By way of example, if you have something stored on your DVR, when you watch it you are not likely to simultaneously use a laptop or browse a newspaper. That sort of behavior is more likely to occur in general schedule-based viewing when programming is more often that of convenience or default (in DVR households) rather than purposeful choice. After all, DVR viewing is simply the modern version of appointment viewing, with the appointment set by the consumer (it's also frequently social viewing).
At this year's ARF Audience Measurement 3.0 Conference I gave a glimpse of some findings from the Media Acceleration Project (that was funded and supported by Time Warner, Pepsico and one other major advertiser) which investigated the likely impact on media behavior of the increased penetration of many digital devices into the home. One of the most interesting findings, to my mind, was the lessening of concurrent media exposure (or media multi-tasking) as a result of the increase in consumer control over the selection and timing of access to content.
I'm inclined to agree with the VSS findings that there is not an infinitely increasing amount of time that we will spend with media -- and also that any decline in time spent with media may be due in part to efficiency gains. However, I think that the impact of increased consumer control of concurrent media exposure will also contribute to an overall drop in media time.
This of course is not necessarily a bad thing. Many people have been concerned by the challenges of how to define and deal with the impact of concurrent media exposure (dilution of attention, etc.). Whereas concurrent media exposure isn't going away anytime soon - and never will - maybe less of it will lead to a greater proportion of what could be thought of a quality time spent with media, rather than simply quantity.
And surely quality is more important than quantity now that so much time is spent with media. After all, isn't that what the whole engagement discussion is about?