According to the author of Learning to Teach Through Video we cannot process 3 different instructional mediums at the same time. Thus, a screen cast needs narration and picture or text and a picture, but not narration, text and a picture.
How does that affect our accessibility push? I guess it means that one section should be a base and the other sections should be optional. In other words, the picture (for example) would be the base and individuals can choose between narration and text.