Even Scott Aaronson doesn't really know what they will be good for. Something something quantum simulation something something material science, perhaps. Yet nothing concrete, no example of a problem that they would actually solve. Apart from cracking RSA, which is the opposite of useful.
It seems the whole quantum computer thing only made people like him excited because it's a different kind of computer, not because there is any strong evidence that it will be practically useful.
It reminds me of nuclear fusion: Sounds cool, but it is highly doubtful whether it could ever compete economically with nuclear fission.
They did a test of pretty much the only thing it can do.
This is an extremely limited system, different in capability but in the grand scheme of things conceptually similar to implementing the quantum equivalent of a single binary digit.
They never talk about what this computer is actually doing.