What about transpositions into and out of book theory for example in the diagram, a series of ridiculous moves that the computer can and should be screaming at both patzers transposes into a position from opening theory. should the player(s) who get it analyzed lose the instructional value.
Well dang -- that does make sense. 
With that in mind, perhaps it could use notes like "out of book" and "transposition to book line" rather than the redundant notes. I don't like the redundancy or it's messy appearance, especially as I take the time to import my games into Chessbase and merge the computer's notes with my own.
The workaround of editing the PGN file before importing into Chessbase is not terribly onerous, but this suggestion would seem to be easy to implement (since the computer already seems to recognize when the relevant events occur).
Also, if Chess.com does make a change, I wanted to note that the current PGN also adds an extra space after the ""{(Book Move)}" string, and appears to add line unnecessary line feeds (unless my processing is inserting line feeds).
Instead of redundantly posting "book move" after every single book move in the opening, why not just indicate "book move" at the last "book move" with the understanding that all previous moves were also "book moves?" Maybe I'm being overly anal retentive here, but seeing "book move" a dozen or more times esp when I already know they're book moves doesn't help my understanding of the game, and anyone who doesn't know which are/aren't book moves can surely understand that every move preceeding the notation "book move" was also a book move.