Re-sync legacy epic hierarchy

We found that a very small amount of instances (~1% of self managed instances from which we receive service pings and use epics) do not have all records for legacy epics and work items in sync (investigation).

We're not able to figure out the root cause as we're not seeing this issue on SaaS, and 99% of the other instances work correctly.

For this reason we have the epic Handle out of sync issues between legacy epic a... (&15435 - closed) which will help us with adding more constraints to the database to ensure consistency.

However, to also help these instances to get to a state where data is in sync again without manual interactions, we decided to re-sync the data in a batched background migration in the %17.5 release to prepare them for a smooth upgrade to the %17.6 release.

We want to:

1. Check if the instance is affected (check if any data is not synced correctly)
2. If there is a syncing problem: Re-run the backfills we already created for our initial backfilling:
   1. Backfill `work_item_parent_links` based on epic... (&12741 - closed)
   2. Backfill `work_item_parent_links` based on `epi... (&12742 - closed)