Relative Efficiency of Maximum Likelihood and Other Estimators in a Nonlinear Regression Model with Small Measurement Errors