Hi @Experimenter ,
you are right: my example is simplified. I the real table I am facing 5-8 types of just “diameters”, it all adds up to dozens of measuring value columns in total. So my first approach with expressions/rules was not up to the task; I had to endlessly combine MISSING/ NOT MISSING arguments which is not practical.
To answer your last comment: as mentioned above, I have different, very specific measures, like diameters d, d1, d2, D1, D2, which have specific meanings in the industry. If still two different types of product share the same set of features they could still be further differentiated by the product name/description or other parameters, but technically they would share the same product model, which is as far as my task of implementing the solution goes.