How well does model fit real data? (II)
Good for single asymmetric peaks
Problems for shoulder and double-peak behaviour
Especially bad at fitting region between peaks