I don't understand the coding well enough to know whether the mismatch could ever matter. I was trying to get fetch annotations to work and discovered that it wouldn't fetch for some of my books that still had old names on the Kindle (sent to older Kindles by much older versions of Calibre) - names with the "The" at the beginning of the title, for example, and with a number appended to the end of the name; I had to resend those and either delete or rename the mpb file to match the new name. I have no idea why this mismatch between the current settings for title sort and file naming and the file names on the Kindle (created by older versions of Calibre) created a problem, so someone else could say better than I whether the mismatch I pointed out could ever make a difference for Calibre purposes. You know better than I; it just seemed to me, as a very amateur coder, like it's safer in the long run for this sort of thing to match or at least to understand why it's not (is it doing the same thing for other devices?)
It doesn't matter to Kindle right now of course - though I guess it's possible it could at some point in the future.