Comment text appears below:
The topic is very interesting and relevant, however I think the author has confused the concept of poor temporal resolution with a type of non-linear motion distortion called "judder". A film or sequence of still images taken with a relatively large time between each exposure produces poor temporal resolution. The film rate of 1/24th of a second can be considered a large time in this context compared with 1/60th of a second for example. This poor temporal resolution can be seen as a jerkiness in motion, especially noticeable when panning cameras or zooming.

Judder on the other hand is a type of distortion to the motion that occurs when the number of frames of the source are not evenly divisible by the number of frames used in the display for the same period of time. For example for a film two frames are captured in 1/12th of a second and when this is displayed at 60 fps, it is necessary to display 5 frames in the same 1/12th of a second. 2 frames won't go evenly into 5 so there must be a 3-2 "pull-down" sequence. One frame (which was originally captured in 1/24th of a second) is now displayed (at 60fps) over the time interval of 3 frames (1/20th of a second) and the other frame is displayed over 2 frames (1/30th of a second). The judder is the time distortion (expansion from 1/24th of a second to 1/20th or contraction to 1/30th of a second) of the motion in the scene.

This judder is visible as a non-linear or uneven jerkiness. Judder is quite distinct from the linear or even jerkiness caused by the poor temporal resolution. Removal (or prevention) of the judder distortion can be done in only two ways: (1) prevent it entirely by using a display frame rate that is an even multiple of the source material such as 48 fps, 96fps, 120fps, 240fps etc. or (2) generate at least one interpolated frame as the in-between frame(s) in the uneven sequence to "blur" the motion across the uneven number of displayed frames.

The first solution completely prevents the judder, but does nothing to minimize the affects of the poor temporal resolution. Depending upon the sophistication of the algorithm used in the second solution, it may only minimize the judder or it may also minimize the affects of the poor temporal resolution. Obviously the best solution is to combine the prevention of the judder in the first place (direct conversion to an even multiple) along with the generation of interpolated in-between frames.

I would submit that there is no real problem with the use or delivery of 24p video content. In fact it does present a more pure transmission of the film-based source material and is the only way to achieve the best solution. In this case, the author has simply identified that his BD player does a more effective job of generating the interpolated frames needed to be shown on his 60 fps display than the algorithm used in his projector.

If a display can take the 24p video from the source and display it natively at 120 fps or 240 fps it has the advantage of never having to compensate for the judder distortion and the algorithm can more easily focus on minimizing the poor temporal resolution. Of course your mileage may vary depending upon the sophistication of the motion interpolation algorithms in the display. Strangely enough, not all 120Hz displays can even accept the 24fps inputs.

On the other hand if your display natively displays at an even multiple of 24 such as 120 fps or 240 fps and the output of the BD player is 60 fps from the source material at 24 fps, you are splitting responsibilities between two non-cooperative components. The BD player is trying to generate the interpolated frames to hide the motion judder and the display is trying to "undo" the motion judder at the same time it is trying to smooth the motion-flow to compensate for the poor temporal resolution. It is certainly better to let one device do all the work, than having them fighting each other.