I agree. The recent grey-scale thread you reference is a rather extreme example in that there was a fundamental misunderstanding of the difference between subject luminance/reflectance and paper reflectance. It is a common error people make when reading Adams etc. Often people look at the page that shows the patches of grey tones from black to white, and they assume that is the range of reflectances they will be metering.
But aside from such glaring errors, even if one knows what the results should be, the actual design of an appropriate test is often difficult. There are limitations to what kinds of situations we can easily create under controlled conditions, counterintuitive as this may seem at first. And one must be careful not to introduce new sources of error when solving a given experimental problem. This is particularly true when attempting in-camera tests. I was on the right track with the test I originally posted to start this thread, but it was still of limited value because of the inherent constraints.
For example, a major constraint in a controlled test for flare is the luminance range we can generate for the target without creating measurement problems.