film editing

You have it about right. I can't remember that actual specifics, but the "jump" at the beginning and end of the cross-dissolve is due to negative cutting, and poor technique at that back in the day I think (cement splicing?)

Whether the dissolve is made from original negative or a dupe IN (duplicate internegative), I don't know or recall, but all the shots required for a fade out are gathered onto roll A (for example) and put onto a contact printer. The printer is programmed to have full exposure on frame 1 of each shot, and to fade out for, say, 48 frames—like your example.

Roll A is removed from the contact printer (all of this done in the dark) and roll B is threaded up with all the shots that are supposed to fade in. Contact printers can run forwards or backwards, but however it's set up, roll B needs to be synchronized perfectly.

Both rolls A and B were pressed up against roll C, which is now processed as either a positive or a negative, depending on whether rolls A/B were positive images or negative images.

Either way, the cross-dissolve negatives are spliced into the master negative, and voila. Over time, the cement becomes crusty or breaks, and the negatives "jump" when run through the telecine to create the movie we see.

Also interesting to note is that the contact printer could've been set improperly and so the exposure might be off by a hair of an EV, so that makes the cuts more noticeable. Also, different film stocks were used between camera negative and contact printing negatives, so that might cause a difference in color temperature and grain, especially over time. So on and so forth. There's a multitude of reasons why these dissolves (and sometimes even title sequences) become noticeable.

Almost everything is done in a computer these days and lasered back onto film, seamlessly. even IMAX films are recorded and lasered out now.

Actually, the problem is the Bausch & Lomb CinemaScope anamorphic lenses, which were noted even then for their lack of sharpness.

As lens makers other than Bausch & Lomb got into the market throughout the remainder of the 1950's, things improved - most notably with the superior lenses from rival Panavision, which eventually demanded the screen credit "Filmed in Panavision."

The jump noticed before dissolves IS caused by the editing technique previously mentioned, but the optical effect (or defect) is more noticeable when the image is optically re-formatted by the anamorphic lens.

Look for this particularly in non-Fox CinemaScope films from 1953 to about 1958; in Fox films from 1953 to about 1961. By the time the Panavision credit begins appearing (about 1960-62) the effect is gone. Most Fox pictures from about 1961 used Panavision lenses but retained the CinemaScope credit until 1966 or 1967.