The coefficient (k) of the Brutsaert-Niber equation (-dQ/dt = kQ^α, Q being discharge at time t) cannot provide information on streamflow dynamics independently because its unit depends on the exponent α. One way to address this challenge is to compute k after fixing α at α median, which may involve large fitting errors. A recent study has therefore adopted a method to decorrelate k from α by resealing Q and demonstrated that the decorrelated coefficients hold useful information on streamflow dynamics. Here, I argue that decorrelation is not dissociation and propose a framework to evaluate the parameter estimation methods quantitatively. Analysis of real as well as synthetic recession curves suggests that under no circumstance the decorrelation method is superior to the fixed exponent method. To conclude, there seems to be no rational solution to the Brutsaert-Niber parameter association problem.