I actually have a pretty good idea about my full reads. I went back into KU about 10 days before KU2 started with one series (3 books). I saw their number of borrows per day over that period, and then, once KU2 began, I saw the page reads. The page reads divided by the number of KENP/book gave me . . . right about the same numbers as for the previous week. That and reviews (I very rarely get "DNF" reviews) told me that most readers read my books all the way through. They may hate a character, they may write a dissertation about it, but they read the book.
I've also been able to see, because of being out of, then in KU, how much cannibalization KU causes. In my case--a LOT. At least 50% of my sales. On the other hand, I get borrows that would NOT otherwise be sales. Net gain.
I don't track because the numbers of "sales" or "reads" are kind of meaningless due to money. For example, I did track the first 3 months of two releases, one tradpubbed and one indie, last summer, and another set of two releases, also one tradpubbed and one indie, last winter. Each of these was two books published about a month apart. Overall, I found I had 40% more borrows/sales (both counted the same--earning the same) on the tradpubbed books, but made 60% more money on the indie books. That has changed over time, as the tradpubbed books get more push.
What matters most to me? How much I make on a book. Besides, tracking is kind of meaningless. I can tell, broad-brush, everything I need to know. For example: I have one series that is sort of "evergreen" without promotion, and also performs the best in audio. I have one series that does consistently well because it gets pretty consistent promotion. I do very well in German. My audio drops precipitously when I haven't had something out for 3+ months, and I need to fix that. Etc.
I used to do corporate stuff where I looked at figures a LOT. I found that unless I were drilling way down, the macro level tended to give me the most information. Actually, I probably get my best information from reviews--the qualitative--rather than the quantitative, beyond general level of rank/money.
Besides, what I really care about is that I make enough so I don't have to get a real job. My income has actually been surprisingly consistent over the past 3 years, much as I always fear it will fall off a cliff. So I don't track.