I don’t know about the content evaluation part. I guess there is none.
As for the soft 404: two things came to mind
URL inspect does JavaScript rendering. But there might be delays or issues when the real crawl happens. If JavaScript injects important information into the page and it fails, this might explain it.
Another thing I remember is an edge case: If the page resembles an error page, because it is an article about error pages, it might be mistaken for one.