I wasn't thinking so much about breeder ratings but individual garden ratings. My thinking being that since the development and severity of rust depends entirely on environmental conditions (assuming the presence of spores and a susceptible plant) then if a cultivar doesn't get rust in an individual garden we do not know if it was not susceptible or whether the environmental conditions were not conducive. On the other hand if it does get rust then a single rating shows it is susceptible to at least some degree even if environmental conditions had an impact on the severity.
We found examples of that in the survey you mentioned - there were some cultivars with very inconsistent ratings from different gardens. There are even going to be places in an individual garden where the microclimate differs, so one may find a given cultivar appears to vary in susceptibility depending on where it is planted. So that's why to my mind a susceptible rating is more useful than a single resistance rating. Now if several people find a cultivar to be resistant in different gardens it would give me more confidence in that rating.
Of course we now know there are different races of daylily rust which further complicates matters since a cultivar may be less or more susceptible/resistant depending which variant of Puccinia hemerocallidis it encounters.
Someone correct me if I'm wrong but I believe ATP has two rust rating systems in the database. One is from the research articles listed in the explanatory article (the one that gives a numerical rating), and the other is open to amendment by individual ATP members. The latter means that someone can rate a plant as "shows susceptibilty" and then somone else can afterwards change it to "shows resistance".
