[ProDigit]

Sorry for barging in, but, am I correct in saying that the 'azw3' file format is the new Amazon format?

It seems to be able to hold more information, formatting options etc...
But how's it's garbage collection? Do the files contain a lot of garbage info?

The fact that a book becomes 4-6MB in size, may have to do with Calibre.
It is an automated program, which still has many errors.
There's nothing like hand-creating a clean HTML0 file, and converting that with some app to the desired format. It beats converting anything else via Calibre.

Calibre may make it appear as if the file loads well, but underneath may be a lot of code hidden, that the ebook reader never will be able to use; broken code, or code that's incorrectly interpreted, and thus incorrectly encoded.

I've said this in the past, how important it is to keep a book formatting very basic (and get rid of class and templates).
A lot of people used not to agree with me, because back in the days ebook readers had SD card slots. Nowadays they have 1 or 2 GB of internal memory, and no way of expanding it, and the importance of keeping files clean and small becomes apparent again! (not to mention smaller files cause less crashes, low memory warnings, and may increase battery life, and book responsiveness; although for regular books you won't notice lots of responsiveness, for books with a lot of chapters, like bibles, or concordances, or dictionaries, responsiveness may be affected).