As a book historian (among other things) who works primarily on the late antique codex — that is, the “book as we know it,” with pages and covers, in the first few centuries of its emergence as a significant alternative to, then its displacement of, the book written on papyrus rolls — I found this very stimulating. (And yes, I write like an academic. I am one…)
In particular, your reflections on the “boundedness" of the book vs the “unbounded" web led me to the following reflections.
In a strange sense, the “book" in the age of the papyrus roll is like the “book" today — except without the concept of intellectual property and all that follows from that. When a book, in the sense of a work (including an edited collection, however miscellaneous) *had* to be explicitly defined, as “these 8 rolls and their content, but not the ones on either side of them,” their boundedness — and, many would thoughtlessly add, their internal coherence — was a crucial defining feature, just as it is now, in the first decades of the “boundless” web. The late antique book, made up of one codex or several, often had a similar character: thus, for example, Augustine’s remarks, at the beginnings of the books and sections of his gigantic but bounded City of God, on each one’s place in the intellectual (and physical) structure of the work.
By contrast, in the long ages of the medieval manuscript codex, the pre-print book par excellence, it seems that the physical boundedness of the codex, ironically, considerably reduced the felt need for “books" to be thus bounded. Without a competing model, codices, always already bounded, came to contain extraordinarily disparate types of content: books, in the sense of works or defined subsections of a work; a subsection or selection from a work, or a canonical collection, often with additional features adapting that material to a new use (I’m thinking here especially of liturgical codices and codices laid out to present a text with its commentaries on a single page); and, most challenging to conceptualize, collections and miscellanies, many of them the result of the physical re-making of a given codex over decades or even centuries. Because the manuscript book *was* inherently bounded, and simultaneously, a unique object — no matter how many individual copies of a given “book" might exist — its users' adaptations, very much including their marginal scribblings, were as much a part of it as was the work, or collection of works, it contained. A messy world…
The way you characterize the “book” seems to me much more a phenomenon of the age of print, though I suggest this only tentatively. Certainly, it is charateristic of the “book" in the age of mechanical reproduction (one could develop quite a tangent here…).
Anyway, this is far too long already: just wanted, by means of a response, to thank you for all the ideas your lovely piece stirred up.