Reading this thread, I get the impression that many pilots aren't happy with the CAD test, but I'm not sure why.
I read through the CAA's paper on minimum color vision requirements for flight crews, and it is quite clear that the CAA wants to pass more pilots, not less. There are at least three objectives that they seem to be pursuing: (1) They want to reliably screen out only those pilots or potential pilots who truly cannot see colors well enough for flying; (2) they want to eliminate any possibility of cheating or inaccurate testing insofar as possible; and (3) they want a test that will actually quantify a person's degree of color deficiency, instead of just pass/fail.
The CAD test is exceptional in that it gives an indication of the degree of color deficiency. It's not just pass/fail. If a person doesn't have normal color vision, the CAD test identifies the abnormality AND gives an indication of severity, which most other color vision tests do not do. This allows the CAA to set thresholds for each type of color deficiency; anyone below a certain threshold cannot fly, whereas anyone above it can.
With other tests, there's no way to know how severe a person's deficiency is, which implies that many people tested must be rejected even though they might see well enough for flying, simply because there's no way to know how bad their deficiency is, so the worst must be assumed.
The CAD test apparently grades test subjects with standardized deviations from normal vision. By carefully setting the threshold deviations that are acceptable for pilots, it should be possible to pass more pilots who might have failed in other tests. Overall, about 35% of people with deficient color vision can still see colors well enough to fly, and the goal is to make sure that anyone with color vision that's good enough can get certified.
In contrast, something like Ishihara or Dvorine plates give only a very coarse indication of color deficiency. Miss a plate or two and you fail. In reality, your color vision might be good enough to fly, but since these tests do not provide detailed quantification of a color deficiency, one must err on the side of safety and reject many who might actually be qualified.
Another, lesser goal of the CAA is apparently to eliminate cheating and inaccuracy. Cheating would be something like memorizing the Ishihara plates, and inaccuracy could result from improper lighting, from holding the plates the wrong way, or from shifting colors in the ink (printed inks shift over time, and since the differences are so subtle in color vision tests, only a brand-new set of plates kept in perfect condition can really be trusted).
Color deficiencies are congenital (in the context of this discussion) and do not get worse over time. Age has very little effect on color vision, and someone who has normal or good color vision won't get significantly worse over time. However, there are some acquired conditions that can impact color vision, so retesting is designed to catch these, particularly in persons who already had a deficiency.
A person with normal color vision will always pass the tests—the false positive rate for tests such as Ishihara, Dvorine, and CAD is effectively zero. So a person who fails the tests under slightly less than optimal conditions almost certainly has some sort of anomaly in his color vision … the only question is how bad that anomaly is, and the CAD is supposed to be able to measure that.
The most important color distinction task in flying is the PAPI, and I note that the CAA's paper appears to use this to establish a threshold of CV anomaly for granting certifications and designing test criteria. It's practically the only situation in which confusion between white and red could cause a serious problem.
It has occurred to me that if the PAPI issue could be resolved, a lot more people with color deficiencies could fly. But that would require a huge, huge effort worldwide to replace these devices with something friendlier to people with CV anomalies, so I suppose that isn't going to happen. And since every pilot is likely to have to use a PAPI sooner or later, during a critical phase of flight, it can't really be allowed to slide.