On the assessment of cardiomechanical function via wearable sensing systems: harnessing population-level patterns and dynamics for robust physiological monitoring