If you're in the extremely large database sphere, you may find that backups running for 26 hours just doesn't work. Especially when restores can take 30 hours or more. Few businesses find this acceptable. So yeah, for very large databases, SAN snapshot & recovery is frequently the answer. BUT, this does mean point in time restores go away and this has lots of implications. This must be discussed with the business and planned for in regards to your recovery point objectives and recovery time objectives. Further, you have to make sure that your SAN snapshot capabilities are compatible with SQL Server. Some of them take this into account. Some don't. You need to test a restore from the snapshot to ensure you're not just corrupting your databases.

This is absolutely not a minor task.

If you're in a cluster, you can't attach the tape directly to one of the machines and expect it to work in a failover situation. If the system is failed over because the server went down, nothing will be accessible. It would have to be attached to a different machine in this situation and accessed remotely (making it even slower).