In section 15.2 on page 378 figure 15.7 is not to scale. The optical rays don't line up.
I conjecture that the diagram was originally drawn without a projector, drawn assuming parallel rays from a far-distant illuminant ... and then the projector was added at the last minute. The distance from the projector to the disk is barely more than one disk diameter. This will result in nearly 2-to-1 variation in the speed of the shadow on the screen. Also the size of the shadow of the disk will be wildly different from what is shown in the diagram. Et cetera.
Suggestions:
One possibility is to dispense with the projector entirely, and indicate incoming parallel light using the same representation as figure 15.6 ... or some other representation. There are many possibilities.
When doing the experiment in real life, the rays will never be perfectly parallel. Therefore the disk should be as close to the screen as possible. The figure should be redrawn accordingly. To make room for it, the mechanical oscillator can be relocated. Its position is not critical, in accordance with the laws of single-point perspective.
OTOH if it is desired to retain the projector, the diagram could be redrawn to use the full width of the page (not just half) for each of the four geometries. This would decrease the nonidealities by an order of magnitude. Rationale: When the nonidealities are only 10% of the story, students can use their imagination to extrapolate to the ideal case ... whereas when the nonidealities are on the order of 100%, this is incomparably harder.
In section 15.2 on page 378 figure 15.7 is not to scale. The optical rays don't line up.
I conjecture that the diagram was originally drawn without a projector, drawn assuming parallel rays from a far-distant illuminant ... and then the projector was added at the last minute. The distance from the projector to the disk is barely more than one disk diameter. This will result in nearly 2-to-1 variation in the speed of the shadow on the screen. Also the size of the shadow of the disk will be wildly different from what is shown in the diagram. Et cetera.
Suggestions: