One big advantage is that the proportionality constant between the measured quantity and the weight of the analyte can be derived from known physical constants. I did a coulometric analysis once, and used Faraday's constant to determine how many electrons were basically flowing - thereby using some stoichiometry and determing concentration of the unknown. So basically, no standardization or calibration is required. Also Coulometric methods are faster and more convenient than volumetric or gravimetric procedures and they are readily adapted to automation.
don't copy my stuff word for word - or I will find and destroy you. and your lab.