Option D is the correct answer because it uses gsutil's built-in hash commands to efficiently validate file integrity without unnecessary data transfers or custom development. Specifically: (1) gsutil -m uploads files efficiently; (2) gsutil hash -c generates CRC32C hashes for on-premises files using Google's standard algorithm; (3) gsutil ls -L retrieves the stored CRC32C hashes from Cloud Storage; (4) comparing these hashes confirms integrity. This approach minimizes cost by avoiding data egress (unlike A and C, which involve downloading files) and effort by leveraging native tools (unlike B, which requires custom Java code). The community discussion strongly supports D (95% consensus), citing Google documentation on gsutil hash and data validation best practices.