Intuitively (i.e. probably not entirely correct) sleeping in a warm bed will use up less calories than a not so warm one, because of less need to keep oneself at the right temperature. But beyond a certain point (or in a hotter climate) you're going to be too warm, needing other body regulating mechanisms to kick in (not sure how much calorific value goes into producing sweat) and perhaps rather than sitting snug, tossing and turning under the covers.
Not to mention other night movements/racing heart arising from dream-induced movements not entirely suprressed by the REM atonia mechanism.
And, indeed, compare with watching TV, where one can find that physical movement and internal functions (including depth of breathing, and often heartrate) has become suppressed, and quite possibly a lot of the usual conscious and semi-conscious thought processes, while the main focus of concentration is narrowed down towards the TV programme. (Although if you're in the habit of snacking in front of the TV, that would definitely skew the results!)
As someone who often listens to recorded radio shows while reclined in bed, with a similar style of aural concentration/submasking-of-input, I might question if I also burn less calories while doing this than when similarly positioned but actively brainstorming some sort of project or other (in the vain hope that I might remember all my clever solutions, or even that I had any, when I awake in the morning!).
I know I go to sleep quicker when listening to a number of my favourite shows (which can be annoying, as I will probably have missed a proportion of the final one) than if my mind is whirring with active ideas and plans and whatnot, so that's one thing I would need to account for in a comparison, as is the fact that I will stir myself to go to the computer and key up the next recording in that set. Apart from not having a whole-body-calorimiter at hand, it's probably not something I could analyse single-handedly, in any case.