Do we have good reasons to suppose this isn't a lie?
Absent that, I would expect the 2-3x on what's made public.
Absent that, I would expect the 2-3x on what's made public.
The carbonite product page broadly lines up with the claimed specs (~5km swath, ~1m resolution). At the claimed orbit (500km), a ~0.5m aperture (see "washing machine sized") is diffraction limited at about 0.7m, so there doesn't appear to be too too much wiggle room.
Side note - carbonite doesn't just do imagery - it does video too (https://www.youtube.com/watch?v=Sf_zd9MvW44).
https://www.sstl.co.uk/space-portfolio/missions-in-build/202...
https://www.sstl.co.uk/getmedia/b38389d7-cb07-4308-944a-a916...
Also this satellite was very cheap at 22 million, the US cutting edge ones are rumoured to cost over a billion (and they're much bigger than washing machine size). So I wouldn't expect it to be a lot better than what's commercially available.
If anything, having only 90cm when could be better is the odd thing.
Seems you found the answer to your own question: it would be odd for it to be not better.
The other reason why it might be a lie is because the exact specs of spy gear is often kept secret. It is very hard to disguise the fact that you have launched a satelite, but comparatively easy to keep it secret what exactly is on it.