I'm researching stretchtext these days, Ted Nelson's 1967 conception of an easy-to-use hypertext.
It seems to me that many well-functioning interactive video texts use a stretchtext concept in some way or another.
In hypertext literature, stretchtext is often described as replacement. In a set overall sequence, the user may replace one node with another that is about the same, only longer (or shorter). A different concept is that of the footnote. Gunnar Liestøl used that in the Kon-Tiki and Vikinger CD-ROMs.
In these videos, footnote markers fly out of the main video pane, becoming animated buttons (micons). Clicking a button plays a video footnote. These do not replace anything, but expand on a sentence spoken in the voice-over.
As I am trying to storyboard some stretchvideo sequences, it occurs to me that footnotes seem to be more functional than replacement. Film is a retrospective art form. Scenes are made by mounting cuts together, and the meaning arising from the montage is understood in retrospect. For a user to perform choices in a video, they will have to react to something they have already seen. If asked if they want to replace what they see right now, they wouldn't know until they've seen it.