What work exactly does Miles include in his description of preparation of a 30,000 word book in two hours?

He specifically excludes dictation. He does include selection of ideas and arrangement. He also says it includes "collection", but I'm supposing that he's taking a larger tranche of cards from a possibly massive collection and collecting only those he needs right now? Certainly the reading, thinking, and collecting work can't be included in this two hours of work.

Does he have a better definition of what he means by collection?

https://www.youtube.com/watch?v=SYj1jneBUQo

Forrest Perry shows part of his note taking and idea development process in his hybrid digital-analog zettelkasten practice. He's read a book and written down some brief fleeting notes on an index card. He then chooses a few key ideas he wants to expand upon, finds the physical index card he's going to link his new idea to, then reviews the relevant portion of the book and writes a draft of a card in his notebook. Once satisfied with it, he transfers his draft from his notebook into Obsidian (ostensibly for search and as a digital back up) where he may also be refining the note further. Finally he writes a final draft of his "permanent" (my framing, not his) note on a physical index card, numbers it with respect to his earlier card, and then (presumably) installs it into his card collection.

In comparison to my own practice, it seems like he's spending a lot of time after-the-fact in reviewing over the original material to write and rewrite an awful lot of material for what seems (at least to me—and perhaps some of it is as a result of lack of interest in the proximal topic), not much substance. For things like this that I've got more direct interest in, I'll usually have a more direct (written) conversation with the text and work out more of the details while reading directly. This saves me from re-contextualizing the author's original words and arguments while I'm making my arguments and writing against the substrate of the author's thoughts. Putting this work in up front is often more productive at least for areas of direct interest. I would suspect that in Perry's case, he was generally interested in the book, but it doesn't impinge on his immediate areas of research and he only got three or four solid ideas out of it as opposed to a dozen or so.

The level of one's conversation with the text will obviously depend on their interest and goals, a topic which is relatively well laid out by Adler & Van Doren (1940).

I recall portions about Socrates deriding "writing" as a mode of expression, but I don't recall specific sections on note taking. What is Ann Blair's referent for this?

The "emerging popularity of taking physical notes" seems not to be in evidence with only one exemplar of a student who lost their notes within the Blair text.

Being willing to capture an idea by spending the time writing it out in full is an incredibly strong indicator that it is actually worth capturing. Often those who use cut and paste or other digital means for their note capture will over-collect because the barrier is low and simple.

More often than not, if one doesn't have some sort of barrier for capturing notes, they will become a burden and ultimately a scrap heap of generally useless ideas.

In the end, experience will eventually dictate one's practice as, over time, one will develop an internal gut feeling of what is really worth collecting and what isn't. Don't let your not having this at the beginning deter you. Collect and process and over time, you'll balance out what is useful.

Frederic Paxson took his note file from Wisconsin to Washington D.C. when he went into war service from 1917-1918, which Earl Pomeroy notes as an indicator of how little burden it was, but he doesn't make any notation about worries about loss or damage during travel, which may have potentially occurred to Paxson, given his practice and the value to him of the collection.

May be worth looking deeper into to see if he had such worries.