Wi-Fi mesh networks have become critical for extending Wi-Fi coverage. Their underlying technologies have evolved from the decentralized IEEE 802.11s standard to the newly introduced EasyMesh, which operates in a centralized manner. However, neither approach can dynamically adapt Access Point (AP) and client association control to optimize throughput performance. Although EasyMesh can steer clients to a mesh AP from a centralized controller based on Wi-Fi signal strength, it overlooks other key factors such as AP load, channel congestion, and backhaul link capacity. In this paper, we propose a practical solution called Practical Dynamic Client Association (PDCA), which offers dynamic client association control over time to achieve max-min fairness in throughput performance while accounting for these critical factors. PDCA employs a heuristic-based approach to determine the optimal AP-client associations and then uses the EasyMesh client steering feature to execute association control. Our prototype demonstrates that PDCA consistently outperforms the default EasyMesh operation, achieving improvements of up to 165.2 % in minimum throughput and 60.0 % in aggregate throughput.