This ability of the female to give birth to "multiple species" seems to me best understood as the two "species" not having yet actually become distinct, since the only meaningful definition of speciation is when two sub-populations of a species have genetically drifted so far apart that they can no longer successfully interbreed and produce fertile young.
During the process of speciation (one species splitting into two) there are going to be various messy half-way stages such as lions and tigers still able to interbreed and so not fully speciated (even if well along, and not going to typically interbreed), horses and donkeys still able to interbreed but producing infertile young (mules), and these ants in this strange state where interbreeding apparently only results in males. It would be cool to be able to speed up the evolutionary timescale to see the process happen, but what we have here is like a still frame from a movie.
Your last sentence correctly points out the frailty of our definition of "species". However, this is not the only time our data has confounded our artificial, if often useful, definition of species boundaries.
I wonder what are the most visually, or structurally, or genetically, different animals that can still interbreed. Things like lions & tigers, polar bears & grizzlies, and zebras & horses, come to mind ... what else ?!
For example, American bison and domesticated cattle can interbreed to produce fertile female beefalos, but the males are sterile. Are domesticated cattle the same species as buffalo?
Then there's ring species: populations of animals where population A can interbreed with populations B and D, but not with C, but C can interbreed with B and D. (often the rings are larger than that). For example, the genus Ensatina salamanders here in California can interbreed with neighboring populations as you go around the mountains, but if you drove one from one side of the central valley to the other it couldn't interbreed. We've mostly decided in that case to call them a bunch of different species, but it's a weird case.
Shit gets even weirder when you leave the animal kingdom. All varieties of pepper will cross pollinate. Bacteria just sort of spread their genetic material to anything that's nearby. Don't even get me started on the absurdity of declaring all the asexually reproducing organisms as being single species individuals.
Basically, a species is a group of animals that has enough of the following characteristics that biologists can agree they're sufficiently different things:
1) They appear distinct from other things
2) They exclusively select mates from their group
3) They exclusively produce fertile offspring with their group
4) They occupy a distinct niche in their ecosystem
5) They are more genetically similar to other members of their group than to other things we consider distinct species
6) Their common ancestor with another group we identify as a species is extinct and considered a different species
7) They really seem like they should be a speciesI hadn't really considered the definition of asexually reproducing species - it seems that things are much more clear cut for ones that sexually reproduce since then we can use the more clear cut "point of no return" definition.
I suppose in cases like beefalos and mules, or these ring species, this "point of no return" comes down to is there any path for to the DNA of these divergent animals to recombine, so a fertile female beefalo (or the occasional fertile female mule) still provides that chance.
It seems that in general it's rare for widely divergent animals like zebras and horses to interbreed in the wild, but apparently western wolf-coyote hybrids are not that uncommon, so it's more than just a theoretical possibility. Who knows, maybe global warming will force polar bears to adapt to warmer climates and increasingly interbreed with grizzlies.
In reality, we first categorized life into species because they either looked different or we found them exclusively in different places, and only centuries later did we attempt to figure out exactly why and how this was the case and reverse engineer some sensible definition onto the pre-existing categories, but it turns out there is no single definition that works universally and has zero exceptions. It's frustrating if you're a language pedant who likes clarity, but a lot of categories and definitions are like this.