I've got this feeling that the best way to start an argument among a group of zoologists as a layperson is to ask "So what's a species anyway? What's the distinction between a species and a genus?".
I hear this all the time, but it pretty much universally stems from people not being up to date on the current state of literature regarding species concepts and speciation, and textbooks and curricula failing to be updated. Basically, this argument was put to rest in 2007, when a landmark paper by Kevin de Queiroz in 2007 (required reading in my lab), pointed out that we actually all agree what a species is: a 'separately evolving metapopulation lineage' (actually he had already basically made this point in 1998 and 2005, but third time's the charm).
The perceived 'argument' is simply around the criteria we use to identify species. Speciation is (usually) a gradual process, over which a series of 'species criteria' become satisfied, e.g. the biological species criterion (inability to interbreed with other metapopulation lineages), or the phenotypic species criterion (phenotypic distinction form other metamopopulation lineages). Different instances of speciation can have such criteria satisfied in a different order, and in some cases, some criteria are never satisfied, or at least, not until much much later (e.g. the inability to interbreed [biological species criterion, one of the hardest to satisfy and test] may be the last criterion to be fulfilled, and in some cases can take millions upon millions of years—just look at cichlid fish, which can, in captivity, sometimes hybridise between South American and African lineages that are >45 million years apart! And don't get me started on the sturddlefish!).
This re-framing allows scientists working on different taxa, for which some criteria might be preferable or more detectable than others, to use those criteria best suited to them to help define/recognise/circumscribe their species. For instance, it is impossible for clonal organisms to have the biological species criterion applied to them—no individual can interbreed with any other, so is every individual a species? If you are debating about the primacy of the 'biological species concept', this will always come up. But if we treat that as just one criterion that is just as good as any other, suddenly that doesn't matter so much.
Now, it is important to remember that species are hypotheses—hypotheses concerning which individuals belong to which metapopulation lineages, and how independent those lineages are from others. Species are real, but our concept of them is based on lines of evidence. They can be confirmed as they can be refuted. When the hypothesis is good, we see it confirmed by multiple lines of evidence; when it is bad, it gets messy, and it can align well with some lines of evidence, but not others. This, too, is exceptionally well suited to this reimagined species concept (what de Queiroz variously calls the 'unified species concept' or the 'generalised lineage species concept') that hinges on a central agreement of what a species is, and a variety of criteria that it can satisfy for us to recognise it. This does a much better job of capturing the biological reality of species and their variation and complexity, than any attempt at using a single 'concept' as we might have in the past.
By the way, de Queiroz also has opinions about subspecies, and his thoughts are what my colleagues and I have adopted: incipient lineages on their way to full species status, but with some criteria still insufficiently satisfied.
As ever, a failure of education to keep up with breakthroughs in the field—a problem ever exacerbated in our age of mass scientific breakthroughs and lagging, underfunded, and often anti-intellectual education systems—continues to breathe life into arguments that have long been put to rest.
As to the question of what a genus is, and how it is distinguished from a species, that one is easy: it is a convenient box into which a group of taxonomists working on the taxon in question decided (and/or agreed) to lump a group of species. This same definition holds true for every rank above species. Unlike species, genera are illusions. Families and other higher ranks, doubly so. They try to tell us something about the evolution and history of a group of species that sit inside them, but they will never do as good a job as a real cladogram. And the decision as to which branch on the cladogram should be given a rank and name? that is totally arbitrary, differs vastly by group, and has a lot to do with tradition and very little to do with consistent scientific logic. After all, after Linnaeus, very very few biologists continued to work on more than one major taxon in a meaningful way. Today, I can count on one hand the number of taxonomists I know who work on two majorly different animal groups, and I don't know a single one who works on different kingdoms. Moreover, most of these higher taxon decisions were taken before cladograms had even been invented, long before Charles Darwin was born—and even today, many taxa have only fragmentary, morphological cladograms drawn for them, and no genetic sequences that would allow a more 'objective' placement of rank names.
This has a very important consequence: you must NEVER try to compare higher taxa, especially across phyla. An insect genus is incomparable with a rodent genus is fundamentally different from an orchid genus is not even in the same universe as a bacterium genus. Even though they are both frog families, it is never informative to say ' The family Microhylidae (768 species) has 1.7 times as many species as Ranidae (457 species)'—Microhylidae is twice as old as Ranidae! The people who work on the group matter more than anything. Just look at anoles. A leading group of people working on anoles flatly refuse to accept division of the family Anoliidae into multiple genera (even though this is easy, has been proposed thoroughly and imho convincingly, and would make it much easier to know how anoles you are talking to are related to one another!), and consequently Anoliidae == Anolis—a genus containing 435 species. A wastebin taxon if ever there was one. The point of higher taxonomy is to be useful. This is not useful.
TL;DR: We basically all agree now what species are, and people who make this into an argument show that they are not up with the times, revealing failures in our education systems. Higher taxa, like genera, are conveniences to help us manage species, but cannot be compared. That's not what they're for, anyway.
