Generate grant expansion entitlement IDs without using resource cache.

[ggreer]

This fixes an error with Azure Databricks and improves performance.

[btipling]

What was the bug? I can't tell from the diff

[ggreer]

The cache wasn't always fully populated, and we'd fail to fetch the group because we're generating the wrong workspace URL for Azure Databricks instances: https://learn.microsoft.com/en-us/azure/databricks/workspace/workspace-details

But we don't need to fetch the group to figure out the entitlement ID to add the grant expansion annotation. We just need to know the group ID and the parent resource ID (since groups can belong to accounts or workspaces).