Posting the HN link here because the pgrust author attempts to downplay and explain away the code as various internal optimization artifacts.

They admit that they submitted to ClickBench an incomplete, private build specifically created for the purpose (and including files hardcoded to the Clickbench test names), which could only interpret a limited subset of SQL (just enough to pass the tests). They said it was just overeagerness.

Particularly, they do not seem to think this counts as cheating, because the old engine was still in there technically even though it was disabled:

Once our new JIT compiler supported enough SQL to handle the ClickBench queries, we decided to submit it to ClickBench. AFAIK, ClickBench has no rules about how much SQL you need to support. On top of that, we did ship that we did ship 0.4-preview with our old query engine. It would have been a two-line change to fall back to it and support a much broader set of queries outside what our new engine could support.