Wrist-worn surface electromyography (sEMG) presents a promising approach for unobtrusive gesture recognition in consumer wearables, but the absence of standardized electrode configurations poses a challenge for optimal system design. This study systematically investigates the relationship between electrode placement, channel selection, and machine learning performance by evaluating four classifiers which are linear discriminant analysis (LDA), k nearest neighborhood (KNN), support vector machine (SVM), and random forest (RF) with time-domain and autoregressive features. Our results reveal a consistent optimal 4-electrode symmetric configuration (positions 1, 2, 4, 5) across all models, achieving 62.1% cross-session accuracy (LDA) while minimizing hardware complexity and cost. Notably, the optimal channel selection is model-independent, suggesting generalizability across classifiers. The system demonstrates strong cross-session robustness, with less than 1% average accuracy degradation over 21 days for seven gross hand gestures encompassing common daily activities. Furthermore, we introduce a novel gesture-specific electrode optimization principle to enhance performance. These findings offer practical design guidelines for balancing accuracy and efficiency in sEMG-based wearable devices.