I have the highest regard for this history, or I would not have both hardcover and Kindle versions of it. These notes concern the transition to the Kindle format, as experienced on the original Paperwhite.
Navigation is mostly excellent, as it should be for a work which will be used as a work of reference as well as one of which substantial portions will be read from beginning to end: there is a table of contents with the major divisions of each of the Four Parts (originally published separately), and at the start of each Part there is its own more detailed contents section. Cross-references and footnotes are hyperlinked. The only improvement to navigation would be if the page references in the Index were hyperlinked too.
The smaller typeface used in my printed volumes for whole paragraphs of quotation is missing, except in Part Four. Thankfully, in the other Parts there is extra white space before and (usually) after the quoted passage. This white space, plus the almost inevitable reference to the cited work, ensures that in practice one can work out where the quotation ends. However, the conventional layout is there for a reason, and we should not have to deduce where a quotation begins and ends.
Some "tabular" textual layouts (which occur infrequently) are not really satisfactory. So two columns of text (e.g. parallel syllogisms, or the square showing "Intellect, Will, Sensation, Desire" at about location 4670) get run together in a way that is frankly meaningless.
There are some strange and arbitrary changes to the printed text: at one point the transliteration of the Greek for "the now" (an italicised "to nun") loses the italics for the article "to". At another, on motion, there are apostrophes representing the primes in p' and p", which is not pretty but suffices. Then suddenly they disappear and what should read « p' to p" » now reads on the Kindle « p to p ». Again, for some reason on the Kindle a space is introduced into the term "not-p" so that it becomes "not -p" and the reader stumbles briefly over whether the hyphen represents the ¬ sign for negation (location about 3138).
There are no illustrations in the Kindle version. The map, as usual on my Paperwhite, is useless.
The page numbers are continuous, doubtless representing those of the one-volume edition of the work rather than of the separately published volumes. The frequent headings (hyperlinked as already mentioned), rather than the page numbers, allow one quite easily to relate the Kindle text to one of the printed editions.
Greek letters get very uneven treatment in the Kindle version: sometimes a proper, scalable Greek letter is used; sometimes a Roman letter equivalent based on the letter's appearance (e.g. Zeta and Zed/Zee); sometimes a Roman letter equivalent based on the letter's position the respective alphabets (e.g. Zeta and F); and sometimes a non-scalable graphic. (A lower-case Greek Phi given the latter treatment looks rather like a smudge on the screen.) Examples follow - all from Volume One, where there are frequent references to Aristotle's "Metaphysics":
Book Delta of the "Metaphysics" is referred to in the print version using the Greek capital letter (Δ), but in the Kindle it appears as book "D". You would expect Book Zeta to cause no problems (capital Zeta (Ζ) and the modern equivalent (Z) being effectively identical) - but sometimes it becomes the capital letter F.
In "Aristotle on Science and Illusion", references to book Gamma use a proper Greek letter Γ (capital Gamma).
In "Essence and Quiddity", references to Book Zeta are to Book F (e.g. location 4112 approx); those to Delta start out by being to D, which then revert to a genuine scalable character Δ (Delta) for the reference to 7.1017a22.
In "Being and Existence", references to book Delta are given as if to book D, and references to book Zeta are to book F. Beta becomes B, and Eta correctly looks like our H. Book Gamma is referred to with mini-picture Γ (capital Gamma) (location 4286 approx).
There is a Kindle typo in the Bibliography for Aristotle (Chapter 2), where the edition of what is called "Metaphysics M and H" should refer to "Metaphysics M and N".
Readers wanting to follow up references to Aristotle will have to be guided by prior knowledge or by the print edition.
As noticed already, footnotes are hyperlinked. The question is whether the links are accurate. A few soundings suggest that mostly they take you perfectly to the required place. In one place (Part 4, Ch 8, n 1) the print version directs you, correctly, to some sections on medieval ethics, while the Kindle hyperlink takes you to one section earlier (to Abelard rather than Aquinas); in another, (Part 4, Ch 11, n 4), the print version directs you to the starting page of the section on "Aristotle's Political Theory", while the Kindle takes you to a more precise location within that section, a few physical pages on.
Reading these pages on the Kindle is generally a delight, but it would be good to see the noticed imperfections corrected.