Structure in Scalar becomes especially powerful when it's used not just as a way to organize content in a publication, but as a way to model theoretical relationships. For example, let's say you're writing about three characters in a movie and the types of shots the director uses to frame them. While in a traditional book you would typically expound on those relationships in prose, in Scalar you could create a page for each character, a page for each shot type, and then tag the characters with the shot types. Now, those connections have become navigable pathways within your book to which you can add illustrative media and explanatory text. Using the tag visualization, you could even automatically generate an interactive diagram of the relationships between the characters and shot types, which could then be used as an illustration in the book.
Scalar content comes in two flavors: pages (i.e. text, specifically HTML text) and media. Anything in a Scalar book by necessity falls into one of these two categories.
The relationships Scalar can make between content also come in two flavors: whole-whole relationships (links between one whole piece of content and another) and whole-part relationships (links between one whole piece of content and part of another).