The auditor-general's report into Health Benefits Ltd came out last week, and it is curiously thin on detail around key components of the scrapped health sector agency's efforts that were meant to save $700 million over five years.
They're the IT solutions for HBL that were supposed to underpin the mammoth, $700 million savings to be achieved over five years.
Take the Finance, Procurement and Supply Chain programme or FPSC for example. This was meant to be done and dusted by November last year, and was originally costed at $68.3 million.
However, the FPSC wasn't ready by March 2015 despite the budget being upped to a whopping $120 million for a pared-down version - which is now said to be ready in 2018.
The FPSC was set to use a Oracle financial system developed with HealthAlliance that appear to have been deployed in 2011, but "this could not be done to the extent originally forecast" the report said.
If you read through the Auditor-General's 26-page report in the hope of learning from it why the Oracle financial system could not be used, you'll be disappointed.
Said Oracle financial system saw $1.9 million being spent on it between June 2014 and March 2015 for continued development but who was given the money to develop what exactly? This is not mentioned.
Using the Oracle system seems an expensive proposition. Between June 2014 and March 2015 "technology operating and licensing costs (such as Oracle licensing costs)" racked up $9.3 million in fees.
Hutt Valley district health board started using a limited version of financial management information system in the FPSC, which "generated operating costs". This included Oracle "application license support and maintenance fees" so high they created an $8 million funding gap, the auditor-general wrote.
If using the Oracle solution cost that much money unexpectedly, did HBL at least look around for a less expensive solution?
Global IT giant IBM also had a finger in the HBL pie, with the National Infrastructure Platform (NIP) that the two announced in February this year.
NIP is not a good acronym, but it is meant to migrate DHB applications and IT systems spread out in forty data centres (apparently) around the country, into two IBM facilities, one in Auckland, the other in Christchurch.
How much is the NIP costing the taxpayer, and will it be delivered on time - if it's continued?
This was said to save $23.9 million over a decade, a figure that HBL in May thought could be as high as $50 million.
It's not clear how much NIP would cost HBL, but veteran IT industry publication Computerworld wrote in July that IBM had "allegedly failed in its bid to deliver on its first contractual deadline" with the Crown health company which, if correct, would surely jeopardise the estimated savings for the DHBs?
Unfortunately, the auditor-general's report doesn't appear to cover HBL beyond March 30 this year, well before the Computerworld story was published.
Even then, Labour's Annette King asked the auditor-general in November 2014 "to look into the performance" of HBL, why it was wound down, what it had cost the health sector, and the benefits it had achieved.
Oddly enough, the IBM - HBL deal on NIP goes completely unmentioned in the report, even though it seems an important component of achieving savings through large-scale data centre consolidation.
How much is the NIP costing the taxpayer, and will it be delivered on time - if it's continued? These seem to be important questions that should be answered.
Given the string of government IT projects gone sideways over the past decades, you'd expect much more scrutiny for them, and transparency so that we can learn from the mistakes and not to repeat them.
When a government IT project blows its budget and is delayed - or not delivered - detailed information as to what went wrong is a must, otherwise there will be plenty more Novopays in the future. And who in their right mind would want that?