A fine example of a ring species, where the species boundaries are somewhat arbitrary.
If you had a dog-free island with lots of game, and introduced some male wolves and female chihuahuas (or vice versa), 100 years later there would be no dogs.
If instead you introduced both sexes of both animals, 100 years later there would be two true-breeding populations. (unless the wolves managed to hunt to chihuahuas to extinction)
Now imagine the same experiment with horses and donkeys, or quaggas and zebras, or any other combination of equids. In this case it is possible for the different species to mate, (I don't know if they do it in the wild or not), but the offspring are almost always infertile. Again, 100 years later, no equids. Again, if you had both sexes of both species, 100 years later you have distinct true-breeding populations.
Why are the equidae different species, but the chihuahua and wolf aren't? Isn't my experimant a good test of specieshood?
This is precisely the fundamental issue. As I understand it, the "theory" of evolution predicts that speciation will occur given enough time and isolation of the populations. The problem is two-fold.
1st, a weak definition of species does not verify the theory. If, as you suggest, we redefine dog breeds as "ring species", where ring species can in fact interbreed, we are assuming our conclusion, which is that given enough time and isolation, those ring species will in fact eventually reach the point where they cannot interbreed. But this is an assumption that is inferred!!!
2nd, as far as I can tell by reading up on the experiments, they have been run on flies (because of the rapid reproduction cycle) and have produced populations which, by preference, choose not to mate with one another. There again, by inference it is assumed that they would eventually drift far enough apart that they could not mate with one another. This may or may not be true, but it is not proof. Remember the joke about "inductive proof" I provided in a previous post illustrating the danger of assuming this form of proof.
The primary reason I criticize the theory is that the weakly worded versions merely apply to breeds within a species while assuming that eventually speciation would occur, and the strongly worded versions do not show absolute proof but again depend upon inference. Now a good theory should be constructed such that one can run an experiment and if one cannot verify the result absolutely, the theory should be chucked out (unless it has some other redeeming traits such as predictability which evolutionary theory lacks because of the time frames involved). Now with evolution no one, as far as I have seen, has really run an experiment under earth natural conditions that shows speciation of isolated populations with a common ancestor population. And that is where the problem comes in.
Let us be perfectly blunt. Science rejects homeopathy and astrology on the grounds that experimental results do not absolutely prove the theory, and those experiments that do support the theory require assumptions and inference. So why do we reject those two theories yet accept evolutionary theory? I argue that the reasons are sociological, cultural, and political, and not scientific.