Regardless of the lumping vs splitting issue (whether you want genus > section > microspecies or just genus > species, where species = section), there is also the issue of misidentification. At the moment all introduced dandelions in North America are identified are either T. officinale (= section Taraxacum) or T. erythrospermum (= section Erythrosperma).
According the the paper discussed in the other thread, just in BC there are 7 introduced sections in the genus. So observations there are being forced into one of 2 species when they could actually be one of 7 sections/species.