@pinestone
Bellow I tried to make my own analysis and interpretation of some images from your movie
https://www.youtube.com/watch?v=klw8y6ksxj0. I assume that in this experiment was used only the blue electron gun, otherwise the transition from one color to another should be in gradient. In these images the transition from one color to another is done directly.
a - observations
--- the shapes are almost the same as learned from school books.
--- the left and right shapes are mirror images of each other.
--- the up and down shapes are mirror images of each other.
--- there are two lines of symmetry in 2D (two planes in 3D), defined by the two vertical an horizontal axes of the magnet.
--- when the magnet is returned upside down, the planes of symmetry are also changed.
b - technical analysis
--- three beams of electrons (or only one) hit always the three (or only one) RGB phosphor vertical strips, from left side to the right side of the screen to create a single line.
--- for a dot pitch of about 0.25mm will result a specific number of RGB vertical stripes according to the physical size of the screen
--- for a 16 inch viewable area will result a number of about 1600 RGB stripes x 1200 lines for a single frame
--- will be required about 16 milliseconds for a single frame, 13.88 microseconds to scan a single RGB line and 8.6 nanoseconds to light a single RGB pixel and change to the next one.
--- 60 frames per second, 16x10^-3s for a single frame, 1.388x10^-5s for a single RGB line, 8.680x10^-9s to light a single RGB pixel.
--- the three beams (or only one) of electrons are deflected in the right or left side of the vertical axis of the magnet, according to which pole is up and down.
--- the three electrons beams (or only one) will be deflected with a greater angle at the vertical axis of the magnet than at a specific distance from it, proportional with the distance followed by the electrons beams (or only one) through the magnetic field until they touch the screen.
--- at the vertical axis of the magnet the distance followed by the three electrons beams (or only one) through the magnetic field of the permanent magnet will be maximum, and at the margin of the CRT's screen the distance will be minimum.
c - shapes analysis
--- the repetition of RGB colors sequence tell me how is the shape of the magnetic field in the plane of that specific electrons beam (horizontal line of scanning).
--- the width of each colored curved strip tell me about the width of the arc segment of the magnetic field. Short width toward the vertical symmetry axis of the magnet and longer width toward the faraway distance from the vertical symmetry axis.
--- the curvature in the vertical plane of the colored strips tell me the magnetic field have a decreasing curvature in vertical plane and all around the magnet toward its magnetic poles.
While for images with the magnet in vertical position, that is, with the magnetic lines of force perpendicular to the scanning direction of the electrons beam, I can understand somehow the resulting colored shapes on the CRT's glass, for the images with the magnet in horizontal position, that is, with the magnetic lines of force parallel to to the scanning direction of the electrons beam, I can't find any explanation for the resulting colored shapes. As you said, what we see here is the result of interactions between the beams of electrons from the three electron guns and the magnetic field of the magnet, which pass inside beyond the glass of the CRT. Of course all that in the case when all three electron guns are working, but according to my experience with delta CRTs, even with a single electron gun the shapes obtained and colors will be the same. Now, you know better if in this movie you had all three electrons guns working or only one.
Now, according to my understanding, the shape obtained with the magnet in vertical position, should be also obtained when it is in horizontal position.
I'll be back soon with my own images and more analysis, because I came into possession of a Sony Trinitron 21" monitor.