Maurice - I understand "the simplest Mendelian case". But I may be misunderstanding what you wrote previously or taking your example too literally ....
I looked up in the Plant Database both cultivars you used in your example (perhaps I was being too literal here?)
White Temptation and Gentle Shepherd both show "sdlg x sdlg" for parentage. I don't know any other information about their genetics other than what the AHS database and the PlantFiles Database wrote. Perhaps you know more. Anyway, I also looked at the children that each cultivar has produced. Most children are light yellow/near-white bloomers.
That is where my confusion about dominant vs. recessive comes in. Why would any white blooming children be considered a recessive gene if both parents are near-white and most of their other children seem to have near-white blooms (depending on what they were crossed with)? What would you consider to be their dominant color?
Perhaps I am just taking your example too literal. I am a visual person, so I like to research and SEE the bloom to get a better idea than just reading a description of the bloom. This is why my confusion.