The SQL resource needs to be dependent upon the disk resources also (among other things). One trick I've used to trouble-shoot is to start the SQL instance via the service manager and treat it as a non-clustered instance. Check the OS and SQL error logs and fix the errors as they present themselves. Once you get the SQL instance running via the service manager, then try starting if from the cluster adminstration tool.

With the disk, network name and ip resources running, I suspect the SQL instance will start via the service manager but not through the cluster admin.