What a good sitemap looks like
An XML sitemap is a list of the pages you want search engines to find. Every address in it is a small promise: this page exists, this is the version to show, and it is worth showing. Search engines trust a sitemap that keeps those promises, and learn to ignore one that does not.
So a good sitemap is short and clean. It lists each page once, at its final address, with https://. It leaves out addresses that redirect, pages that return errors, pages marked noindex, and duplicates with tracking tags. Its lastmod dates are the days pages really changed, not today's date on every line.
How to read the check
The file checks cover every entry: the format, full addresses on your own host, duplicates, the 50,000-address limit and lastmod dates. Then the first 100 addresses are checked one by one, the way a crawler would ask for them. A redirect means the sitemap should list the address it redirects to. An error means the page is gone or broken. Noindex means the page asks not to be indexed, which contradicts the sitemap.
Most site builders write the sitemap for you, so fix problems at the source: in your SEO plugin's settings, by removing old pages, or by updating your own internal links. Then check again, and submit the sitemap in Google Search Console.