It's not new scientific ground but a machine beating a challenging computer science problem unassisted is a big deal. If they can do that then there are a lot of other challenging things they can do.
Small model solves an easy problem; big model solves a challenging problem. I wouldn't call those problems; they are more like invented puzzles. Perfect match for the AI marketing department to "solve".
The marketing speak of some companies goes into effectively fantastical territory. Some claims that are made effectively imply P = NP, which is like bad TV-series-level science fiction, but some people fall for it.