I don't have all the accounting details, but one reason supposedly was that 60% of the money was earmarked to go to the local Australian suppliers.
Unlike many of the other examples you gave, which were built in foreign docks, these French-tech subs were supposed to be built in Australia.
Apparently that also involved building an entirely new construction facility in Osborne, albeit next to some existing ones where older Australian subs had been built; the shiplifting facility was apparently going to be reused though.
Located at the Osborne Techport facility in Adelaide, South Australia, the Submarine Construction Yard (SCY) will be the home of the Future Submarine Program, alongside other maritime defence programs, including the Hunter Class SEA 5000 and Collins Class submarine sustainment programs. The location of the SCY is adjacent to the ANI-operated Common User Facility (CUF) in Osborne, which includes a Shiplift facility which will be used to launch the submarines when they’re completed. [...]
The SCY facility is being designed and built by the managing contractor, Laing O’Rourke, to the specifications developed by Naval Group based on extensive experience in submarine design and build.
Somewhat of an aside: it does look like Australian sub projects have a high cost recently, e.g. the overhauls of the existing 6 Collins class subs (that Australia presently posseses) was put at 6 bn AUD (that's about 4.35bn USD) according to a June 2021 statement of the defense minister Peter Dutton.
One would have to factor in inflation for a better comparison, but in nominal terms that's actually more than what it cost to build those 6 subs in the first place (5.1 billion AUD), according to Wikipedia. The planned Collins refurbishments look pretty extensive though as each sub was slated to spend two years in dry dock. This is also fairly similar to the original build time, approximately 26 months on average for a Collins. It is certainly a lot less than e.g. a Virginia class, which takes some 6-7 years, albeit that is slated to be reduced to 5 as more get produced. The Barracuda/Suffren do seem to exceed the Virginia class in this regard though, seemingly taking 7-12 years. (One would probably also have to consider that the US plans to have 66 Virginias, while France only ordered 6 Barracudas, so there are more incentives and opportunities to speed up production of the former.)
A more recent Sep 2021 press article has put the refurbishment cost of the Collins boats at AUD 9 bn, but it's not too clear what the source of this latter figure is.
So, just as a ballpark estimation, if we consider a triple building time, and twice the number of boats (12 vs 6) compared to the Collins, we'd expect the [cancelled] Attack programme to have cost around 36-54bn AUD. The difference to AUD 90 bn probably were technological premiums etc.
I know this isn't exactly what the question ask, but I found it intersting how the jump from the 50 to 80/90 billion public estimate happened.
Around 2018 when the Australian subsidiary of the French manufacturer ordered the first (test) batches of steel from Australian manufacturers, the cost of the project was still being quoted by the press at around 50 bn AUD.
According to a 2020 news piece, apparently the public was mislead by being told of the lower bound of the goverment's estimates rather than what they really thought it might cost:
it has now been revealed the government budgeted for the project to cost $78.9 billion as far back as October 2015. This was the same month Defence officials told a Senate estimates hearing the out-turn cost was $50 billion.
The disclosure was made by the Department of Finance in response to a question on notice from a parliamentary inquiry into Australia's shipbuilding program.
Opposition defence spokesman Richard Marles said the revelation showed the government had "refused to be upfront about the true cost of the program".
The official response from a Defence spokeswoman was simply that... the 50 bn was a lower bound.
"In the 2016 Defence white paper, the government publicly advised the Future Submarine program would involve an investment of greater than $50 billion. Similarly for the Future Frigate program, the 2017 Naval Shipbuilding Plan outlined its plan would involve an investment of greater than $35 billion, noting for both programs the need at the time to protect the commercial position of the Commonwealth during negotiations."
(Emphasis mine. The cost of the frigates was mentioned because that also went up, albeit in less dramatic fashion only to $45.6 billion by 2018.)
Marcus Hellyer, a senior analyst at the Australian Strategic Policy Institute, said the revelations suggested the estimates were "deliberately sanitised to take the sting out of it".
"One would suspect if they did give a band, the submarines would be $50 billion to $100 billion, and it would be such a huge number it would terrify the Australian public into not wanting to go down that path," he said.