In case of DB disaster recovery timing is quite important and DB restoration takes significant time and for us means these steps:
1) Create new DB from backup (wait for complete)
2) Enable read-only Geo Replication to 2 regions (wait for complete)
3) Take connection string from 2 and change devops configs and schedule config build (wait for complete)
4) Deploy services using config with new connection strings

So step 3 could be done in parallel to waiting for step 2 if we knew connection string ahead of time and save much valuable time in case of disaster