Remove Legacy Cluster Reindexing Service

To support reindexing processes in progress when making the change to support retries, the original service was renamed LegacyReindexingService and kicked off if the Elastic::ReindexingSubtask had an elastic_task record filled out. We need to clean up that logic to remove the legacy service and all related specs in the following files:

• ee/app/services/elastic/legacy_reindexing_service.rb
• ee/app/workers/elastic_cluster_reindexing_cron_worker.rb
• ee/spec/services/elastic/legacy_reindexing_service_spec.rb
• ee/spec/workers/elastic_cluster_reindexing_cron_worker_spec.rb

The following discussion from !55681 (merged) should be addressed:

• @DylanGriffith started a discussion: (+1 comment)

  @terrichu this might be quite tricky to handle but it occurs to me that somebody might be doing a reindex while GitLab is being upgraded and technically it should be safe for them to do so. It seems we are no longer using this elastic_task. Maybe we can use the existence of elastic_task on this top level to decide to use the old version of the code (maybe just create a LegayReindexingService or something? We can then do a follow up to remove that behaviour in %14.0 to be extra safe?