Grid codes increasingly require wind power plants to provide damping for inter-area oscillations. A key performance indicator is the phase shift between grid voltage and the reactive power at point of common coupling, known as the POD-Q response. However, the effective damping contribution is significantly compromised by communication delays between the power plant controller and individual wind turbines, as well as by turbine-level response times. This paper analyzes the impact of communication delays, controller processing delays, and converter response delays on POD-Q performance. Based on this analysis, a WPP communication architecture, a plant-level control strategy, and a practical parameter tuning scheme are proposed. The tuning scheme explicitly compensates for measured round-trip delays using a phase-lead filter. To test and verify the optimized POD-Q function, a hardware-in-the-loop testing platform is developed, integrating the plant controller, communication system emulation, and converter controls. This platform enables comprehensive validation of communication delays, and power oscillation damping capability. A GOOSE-based fast communication system and a plant-level POD control system are deployed in an Italian wind farm. Field tuning tests demonstrate that after optimization, the POD-Q response meets the required open-loop phase shift. For example, at 0.1 Hz, the measured phase shift that reactive power lagging voltage is -150°, which lies well within the required range of -90° to -180° for positive damping under the negative-feedback convention. These results confirm that the optimized POD-Q response complies with the phase shift requirements of the Italian Terna grid code.