Sean: For the paths, that is one way to prevent name/path collisions. You might also want to combine that method with volume mount points so that you don't have multiple databases or log streams from multiple storage groups on the same physical disk. As for activation of an SCR target, that is a manual process that involves using either Exchange 2007's database portability feature, or using Setup with one of the recovery switches: /m:RecoverServer if recovering a standalone server, or /RecoverCMS if recovering a clustered mailbox server.
Moha: There is no fallback in SCR, as fallback is a cluster term and SCR is not a clustered solution. Although you can use SCR to replicate data from a SCC and CCR, SCR does not require clustering.
Yasin: Yes, you can use SCR to replicate data from a clustered mailbox server in a single copy cluster to a Mailbox server in another Site. As for testing, like the work on SP1 itself, testing in labs and in MSIT's environment is still underway, and it does include testing of CCR+SCR and well as standalone Mailbox servers+SCR.