My semi-casual observations of opinions on usage of PVRs/DVRs sets them into three camps:
1) Industry analysts who group all PVRs together
2) Pundits and bloggers who insist that PVR=TiVo
3) Those who like to look at these devices by interface class
The last one seems to be shared by maybe a dozen folks in the world. I am the only one who writes about it I know of (tell me different and I'll link and subscribe to them).
See, TiVo &mdash and all of the Dish PVRs &mdash have a commercial skip feature. In general, they are pretty much immediately reactive, and make sense. Lots, and lots and lots of others (like most given away by CableCos) are not. They are sorta terrible, slow to react, and provide a jumpy fast-forward feature, at best. These are more like using a VCR than anything.
The results of this study seem possiby skewed by this misunderstanding. Or maybe because the study organizers have never seen a PVR (having worked with researchers, not that big a stretch). They claim that most users don't skip commercials (except older folks). This bears out other studies where PVRs aren't scary (to ad-revenue based broadcast businesses) because commercials are still watched.
But I see usage, in my anecdotal experiences, as almost entirely divided along device functionality (or interface class) lines. People with terrible DVRs watch TV like always (static). They periodically watch like they have a VCR (turn on a recorded show and watch it without much pausing or fast-forwarding). People with good PVRs watch TV with almost complete time-shifting, skip intros and title sequences unless they are very engaging, and watch very few commercials (but go back to watch good ones on purpose).
As an aside, I think this is a typical trend of a certain kind of research:
Male, female, age. Demographics can only serve so much. This is where marketing and usability/UX diverge.