Explanation
The typical and worst-case historical response times, and how many concurrent service calls are being placed are two considerations that an integration architect should make when building a scalable integration solution for synchronous callouts to "available to promise" services. These two factors can affect the performance, reliability, and availability of the integration solution, as well as the user experience of the customer checkout process. The integration architect should design the solution to handle high volumes of service calls, optimize the response times, handle errors and timeouts, and avoid hitting governor limits or service quotas.
References: Certification - Integration Architect - Trailhead, [Callout Limits and Limitations], [Integration Patterns and Practices]

Explanation
The system constraint questions that should be considered when designing an integration to send orders from Salesforce to a fulfillment system are related to the performance, reliability, and scalability of the integration.
The latency and idempotency of the integration are important factors that affect these aspects. Therefore, the questions A and D are relevant for the integration design. The question B is related to the business logic of the order validation, which is not a system constraint question. The question C is related to the implementation details of the Outbound Messaging interface, which is not a system constraint question either