"The knuckle rap test reveals a very dead cabinet." The knuckle test is not a reliable test because it is a point excitation of a panel. In reality the well-distributed pressure within the cabinet does the excitation. I well remember a Japanese manufacturer showing me, while on a factory tour (before I joined Harman and became a competitor), that they poured about an inch of concrete into the top of every box so that customers would be impressed by the almost inevitable knuckle test
What matters is the sound radiated from panels, not movement of the panels - and this is measured in spinoramas. Some panel modes allow considerable movement, but the sound radiated from different portions cancel each other. So a knuckle test, or an accelerometer placed at one location cannot reveal what a panel will radiate. It is understanding this that allows good engineers to strategically place internal bracing to eliminate bothersome modes while not spending materials or mass addressing innocent ones. Enclosures do not need to be "inert", only acoustically "quiet". But when marketing gets into the act we get monster massive boxes whether they are needed or not. It is good for imaginations though.
The most serious resonances are most often associated with the drivers.
"I ask because I've never seen a Spin on an angled ceiling speaker.". Nor have I, because that requires a special 2-pi anechoic chamber. One such chamber is currently being built at Harman because of the popularity of in/on-wall/ceiling speakers. However, I am not at all disturbed by this because experienced engineers can anticipate the important performance factors from conventional anechoic data. Resonances are not an issue if high-performing transducers are used and directivity can be measured or inferred from other anechoic data. Subjective evaluations make the final decision, as always.
More research can always be imagined, but having done it for several decades it is clear that it takes competent subjective & objective evaluation facilities ($$$$), skilled/educated staff ($$$$), and time ($$$$). We can always hope for a scientific "sugar daddy" I suppose
But these days money seems not to be drifting in the direction of research that benefits consumers.