You can also probably find it elsewhere if you're not a supporter of your local PBS station, but I'll leave that search up to everyone and their conscience.
Species, taxa, subspecies, hybrids, populations, chronospecies, trunk species, type species.... All words describing a fundamentally non discreet thing. Yet, knowing that, we (anywhere from casual readers to biology taxonomists) get stressed and obsessed when our leaky categorisations leak.
It's a good reminder that we're natural categorizors. Our minds are all about instances and abstract truths. Lassie the dog, coyote the canid, Socrates the philosopher.
In linguistics this problem is called the single other hypothesis and rater popular. In programming and math it's the diamond square problem, e.g. for java's inheritance mechanism, which is countered with generics, traits, typeclasses and so on. Abstract Algebra and Category theory have to deal with this and the corollary from my POV is, that it's just really complicated.
Before genetics, phenotype was based on appearance. Genetic genealogy is a rather young field, so it might take a while for deeper insights to trickle down. Also, it's not completely wrong to base a classification on environment, because a protein to synthesis X from Y is useless, if the environment doesn't provide Y.