There are multiple ways to recreate the index

When troubleshooting problems with Advanced Search, you may get to a point where you have to recreate the index. There are two ways to recreate the index used by Elasticsearch (or Opensearch). The options are:

• zero downtime reindexing AKA cluster reindexing
  • Docs: Zero downtime reindexing for Elasticsearch. This is completely different from the zero downtime upgrades associated with Geo.
  • See the cluster reindexing service.
• reindexing
  • Run gitlab-rake gitlab:elastic:index
  • Deletes the old index first. As a result, should be used as a last resort. There is no index to fall back on while the operation runs.

That supports why we say recreating the index is a last resort: there's nothing to fall back on when you're doing that (the gitlab-rake gitlab:elastic:index method).