The point isn't that ordinary computers can beat Google's QC for some constructed problem. The QC can always be programmed to work on a harder problem.
The issue in this case is that the Google authors misrepresented what was possible on a supercomputer by many orders of magnitude and presented that as an example of supremacy. In addition to solving a useless problem with no application, they basically used FUD to trash their competitor.
If you want t o convince the larger community, always show your competitor the greatest respect and don't handicap them. It's hard to trust Google's team (well, Martinis left) after that.
Do you have a source on this claim of misrepresentation? I thought that at the time of the publication, there was no reason for the Google team to believe that their estimate of the classical difficulty was incorrect. The initial IBM response showed that a novel approach could work, and the reference in this article is novel without a doubt.
The question I'm interested in is that now that we can mimic the computation on a classical computer, it should be possible to see if the Google quantum computer was correct -- previously they had used cross-validation in an attempt to give evidence, if not proof, for the correctness of the QC output, but with this it should be possible to test it directly. That would at least give us an idea of how noisy the qbits were, and whether the noise cancellation actually worked at scale.
The source of the claim is my own authority (used to write codes for supercomputers). They made the most naive analysis of what a supercomputer was capable of. It was an intentional nerfing, done to make their claim look as good as possible. Their failure to anticipate that there were trivial ways to improve the supercomputer approach is a huge problem. It was also pretty amateurish because the supercomputer field has a very long history of being challenged, and responding to that challenge by updating an algorithm to specialize on the problem.
The issue in this case is that the Google authors misrepresented what was possible on a supercomputer by many orders of magnitude and presented that as an example of supremacy. In addition to solving a useless problem with no application, they basically used FUD to trash their competitor.
If you want t o convince the larger community, always show your competitor the greatest respect and don't handicap them. It's hard to trust Google's team (well, Martinis left) after that.