What is a reason for manual creation of a NAT rule?

This is false - R80 and later releases still support and require manual NAT rules for complex or edge-case scenarios that automatic NAT cannot cover.

Hiding RFC1918 addresses behind a public IP for Internet access is the primary use case that Check Point's automatic Hide NAT handles without manual rules.

While service-specific NAT is a valid use case for manual rules, it is not the most definitive reason listed, and option D describes a scenario automatic NAT structurally cannot resolve.

Check Point's automatic NAT uses the gateway's external interface IP as the translated address by default. When the organization uses a public IP that is different from the gateway's interface address - such as a dedicated NAT pool address or an IP assigned by an upstream provider - automatic NAT cannot determine the correct source IP, and a manual NAT rule must explicitly define both the original and translated addresses.