Re: "If you don't understand why, it's not good science"
I'm not actually (formally at least) acquainted with either form of epistemology, as you put it, but I think I know what you mean in your first paragraph.
The problem I had with what was said was with the context. Assuming something like:
1. formulate hypothesis that PV with bumps is more efficient than a smooth one
2. design an experiment to test it, with a smooth control and various different patterns
3. do the experiment and make measurements
4. compare observations to expected results:
4A result support the hypothesis
4B observations that Blu-ray seems to have best efficiency
The context of the statement seems to be about the need to explain 4B with the implication that it's not good science if it's left unexplained. In reality, they just went "that's interesting" and proceeded to try and explain it. If they hadn't, but instead just published the data and results with a note about the interesting Blu-ray results, it wouldn't detract a whit from how good the science of the paper was. They followed up because it interested them (and probably because they wanted to be the first to publish a possible reason "why"). The "it wouldn't be good science if we didn't" argument is spurious in this particular context.
but the man wasn't trying to provide a formal definition of scientific epistemology, so perhaps you should relax a bit
Yes, you're right. It was a throwaway line, but it irked me that it was delivered as a fact about the scientific method. If it were literally true, then we'd end up saying that things like the observation of the Mpemba effect isn't good science because it didn't come with a "why."