As the Tax Office looked shell shocked and analysts wondered at the implications of the rather substantial JobKeeper mistake uncovered on Friday, we are left with more than three million workers suddenly unaccounted for.
They were thought to be receiving the wage subsidy, but turns out they're not. There are 2.9 million workers on JobKeeper, not 6.5 million as we were told last week.
The presumption amid Friday afternoon's head-scratching was that the missing 3.5 million are still employed - and paid from their boss's coin rather than the government's.
If that is the case, it might mean good news - their businesses weren't hit by a 30 per cent downturn in revenue so didn't need to apply.
Or it might mean bad news - the scheme was just too bureaucratic or risky so businesses didn't apply and offloaded staff or working hours instead, but in this case you might have expected those underemployed or unemployed workers to have shown up in labor force data and dole queues.
Assuming the missing 3.5 million are safely tucked inside functioning businesses, there are some puzzling things about the Tax Office error.
One is the sheer mismatch between the expectation that six million people would be on the scheme and the reality that 2.9 million people are. It's a seriously awry forecast. It is puzzling also that the error wasn't picked up earlier.
The Tax Office says it was noticed because the number of payments going out the door didn't match the amount of money being paid out (2.9 million at $750 a week equals $2.2 billion a week; 6.5 million people equals $4.9 billion a week).
It was an impossible moment and everyone was doing their best.Deloitte's Chris Richardson
But you imagine that someone might also have run a makes-sense calculation through the data - like if you buy a piccolo and an almond croissant in Lonsdale Street and you're charged $5, you know someone has missed an item off your bill.
In the JobKeeper case, if 760,000 businesses are claiming the wage subsidy for 6.5 million workers, they're claiming for more than eight workers each. If they're claiming for 2.9 million workers, that's fewer than four workers each. Shouldn't that discrepancy trigger a hang-on-a-minute moment somewhere?
It is, in itself, an instructive number that could use some expert pondering, since even a small cafe in Lonsdale Street probably has more than four workers on the books - and when you consider that big employers like Qantas with thousands of staff are counted in the JobKeeper numbers, an average of four seems low.
Does it imply that a very large number of workers in businesses hit hard by the downturn are not eligible? Or does it demonstrate that businesses on the scheme are mostly very very small indeed?
Analysts do not expect the revision to change the unemployment forecast, but it will change expectations about what happens when JobKeeper ends.
The worry was that turfing six million people off the payment overnight would trigger a massive surge in unemployment and in collapses of businesses being kept alive because the government is paying their wages bill. That risk is still real, but the numbers are vastly reduced.
And while the recount was generally seen as good news, suggesting the lockdown hasn't been as painful as thought it also means there isn't as much government support being pumped into the economy as thought, and there is still considerable confusion to work through.
Chris Richardson, at Deloitte Access Economics, is upbeat and in a forgiving mood. "Let's just take a moment and be happy. There's no point in throwing stones at Treasury or the Tax Office. Everybody involved has been sprinting for months and has slept very little," he said.
While the original forecast of six million workers on the subsidy had seemed high, it came in the same week that Australia's virus cases were multiplying at their fastest, he says.
"It was an impossible moment and everyone was doing their best," Richardson says.
"Why is JobKeeper cheaper [than forecast]? It was bureaucratically difficult to navigate so some people gave up, but the overwhelming reason was simply it was put together when the virus numbers were at their most horrendous and it looked as though Australia was headed to hell in a handbasket."
Brendan Coates, at the Grattan Institute, says there are "just too many conflicting signals to say whether it's unambiguously good news or not".
"The real question now is you've got three million people who are being paid through this scheme rather than six. If that's because the economy's going better than we thought then great, but if it's because a large number of firms didn't want to apply because they didn't want to borrow money to pay their employees then that's a problem."
But if it's good news for the economy, it is less so politically.
"Granted, there will be issues that are raised at the margin on a program of this scale. But let's not lose sight of the forest for the trees here. I mean, this is a program of historic proportions that is helping six million Australians right here right now."
The error was manifestly in the bureaucracy, but the government will wear the political fallout.
The shine has come off and if Morrison was already inclined to shuffle it off centre stage in its June review, he will only be more so now.