I don't exactly "work" in translation, but I have a bachelor's and master's degree in it, so here's my take:
Literally THE golden rule you get drilled into your head when you study translation is that a good, professional translation should preserve all information of the original that is necessary and relevant to the purpose of the translation, while sounding like it was written by a native speaker.
Where the line gets blurred is when you get to the question of "what counts as "relevant" information.
For example, if I have to translate an interview with a doctor about a new, revolutionary cancer treatment, then the way this doctor speaks is almost certainly not relevant to the target audience. What is relevant is that I accurately convey they scientific data and causalities the doctor talks about. People won't care if she spoke with a valley girl accent or if he had a lisp--they care that there's a new revolutionary cancer cure and they want to know how it works, what the risks are, etc.
But let's say I have that same interview in a medical novel and part of the plot twist of that novel is that this doctor seems like a complete valley girl ditz on the surface, or that he has terrible stutters and lisps, and is thus not taken seriously at first, but is actually SO, SO VERY SMART that they found a new, revolutionary treatment for cancer! Then, suddenly, their manner of speech IS relevant, because the reader cares not just for the doctor as the presenter of data (which, this being a novel, has a good chance of being fabrication anyway), but for the doctor as a CHARACTER, and their manner of speech is part of that character. To translate that interview in a flat, typical medical speech way would lose relevant information.
Now, it's important to note that it is not as simple as "has a lisp or valley girl accent in English" = "has a lisp or valley girl accent in language X". First of all, language X probably does not even have a direct equivalent of the valley girl accent, so you first have to understand what kind of associations this accent is supposed to invoke and THEN find an accent in your language that accurately conveys this same information. Same for speech impediments. If a lisp in language A is considered a major flaw, but language B sees it more as a minor annoyance... then you probably want to pick something else to accurately convey the severity of the impact this has on the character's credibility.
So the real question here is, what was the purpose of that interview? Was the fact that Australians are generally more of a masculine culture (if that is true) and that Australian women have a more masculine register of speech, actually relevant to what the interview was about? Was it relevant to what the interview was supposed to convey in Japanese?
If yes, then this Japanification of the translation was a mistake.
If not, then it was in line with the golden rule of translation, because translating the register accurately would have added nothing to the relevant information--in fact, it would likely have distracted the Japanese audience from it, because they would have spent the time they should invest in processing the interview's actual subject matter instead wondering why the subtitles/dub were so alien to what a Japanese person would expect.