I was not faulting the research present per se just asking if all the questions had been asked.
Is there a disease or parasite which can be passed to the offspring in the sperm or eggs?
I don't pretend to know the answer. Did anyone, in their eagerness to provide proof for the evolution of a new species (a modern, rapid example), even look?
I would think that scientists of any worth would welcome the tightening of their attributions of phenomena by the elimination of other possibilities, but I may be simply naive.
As for evidence of a competing hypothesis, if the question has not been asked, the data which would support or refute, should they exist, have probably not even been collected.
Let me say first that your concerns about the "closed-mindedness" of scientists are not without merit. Science has a bad habit of going "LA, LA, LA, YOU'RE WRONG." Case in point: A doctor in 1983 was basically laughed and ridiculed endlessly about her hypothesis on genes that were mobile in the cell. Later, she won the Nobel Prize. Oops!
So I gave your question some thought. Here's what I came up with:
We agree that it's not an extracellular parasite, because that would be visible during sperm and egg fertilization. It's also not bacteria, because (assuming they could enter the sperm or egg cell) they would do what they do best and divide, divide, divide, destroying the fertilized egg long before it differentiated into a tadpole. Even if the bacteria stayed dormant for a long time, it would be visible (and very characteristic) under a light microscope.
So that leaves a virus - one could come up with an interesting hypothesis about a retrovirus like HIV which integrates itself into the host DNA and only gets activated under certain circumstances - there are certainly hypothetical interactions between viral genes and frog genes that could cause problems during differentiation.
But this poses a problem. If I understand correctly, your question rests on the basis that in the absence of the offending parasite/party, northern and southern frogs would have perfectly good offspring. However, in the case of an integrative virus, to remove its influence, we need to remove the problematic genetic material and potentially replace it with compensatory material. But since the frog is completely defined by its DNA, there's no fundamental difference between removing virally inserted DNA and removing DNA that has mutated from other causes - even random mutations - and we're back to the original case that the frogs are effectively two different species because of massive, prohibitive, genetic differences.
Let's assume next that this is caused by a complete, separable virus that infects the adults, travels to the gonads, moves into the egg or sperm of the northern frog species, and can from there become a part of the offspring, activating when it receives certain developmental cues from southern frog's half-genome. However, this poses a couple problems. First, no matter how successful a virus is, unless it is fully integrated into the host life cycle, there are going to be northern frogs without the virus - and some northern/southern matings should be successful. This was not observed (i.e. 100% of the northern/southern hybrids are unsuccessful.) Alternatively, if the virus is a fundamental part of the host cell during all phases of the life cycle, then this discovery would effectively mean a Nobel prize - because it means the virus can replicate at mitosis along with cues from the host cell machinery AND at meiosis - which would be a completely unprecedented discovery and probably win a Nobel prize for showing definitive evidence for the "evolutionary symbiosis" origin theory of cell organelles.
In contrast, it is perfectly precedented (in other "recently" separated species such as stickleback fish) for accrued genetic differences in a species to make their offspring less successful and effectively push them further towards speciation when they meet again.
There are other lines of reasoning and observations (specifically the morphology of the dead tadpoles) that would rule out viral intervention, but this is probably sufficient for our discussion.
Have a good one!