[Sarmat89]

Originally Posted by Hitch

What, exactly, is it that you think DocBook does not have?

Poems, letters, scenic directions, something else...

You think that tagging every element in a book separately, using XML, will somehow be "less complex" than using HTML?

You have to "tag them separately" with any format.

open the first 30 books you see, on Amazon, in Kindle format, and tell me exactly how many of them mention their cover designers.

I didn't mentioned cover designers once.
Of the 17 EPUB books found in a random folder, 15 had names of the cover artists.
All books which had 'maps', had the corresponding artist credited.
Six books also credited interior illustration artists.

use a Heading tag for their chapter heads

See, they can use styles. It is just one step further to mark the major semantic elements.

tagging every single element in their book with things like "colophon" "copyright page" "title page" "half-title page," "parasimple" "parafirstchapter"

Why should they do that DocBook markup in books? I have no idea, too.

Is that so? Which ones?

Out of my head, FBReader and CoolReader. There are more of them.

You don't know because your PITS format is going to leave all the graphical layout to the READER. So, you don't know whether the footnote would be rendered so as to increase the line-height--or not.

There is only one party who knows how to display it in the best way: the reader app. Therefore, it is a natural choice to leave it to it.

wrap XML tags around everything, doing NO formatting whatsoever, and then throw the baby (the layout) out with the bathwater.

The thing is, there are formats which do just this, and they work good. With the right set of HTML, it will work great.
Also, you keep ignoring the fact that you can apply your treasured CSS formatting over the semantic layout, too.
Can you at least formulate, what kind of features of that "manual design" of yours cannot be achieved by the bare XML layout? Item by item.