First decide whether the URL should be searchable. For important pages, investigate server responses, access restrictions, internal links, sitemaps and duplicate-URL volume.
Crawled – currently not indexed describes a different processing stage. Neither is universally more severe; judge affected important pages and persistence.
Noindex must be crawled to be read, so it is not a blanket crawl-budget saving mechanism. Separate index-exclusion goals from crawl control, then request reprocessing for important corrected URLs where appropriate.
