[1] In this paper, the second in a two paper series, we quantitatively compare a detailed theory for type II solar radio bursts with observations and extract the parameters of the associated shocks. We use the techniques and assessment parameters developed and demonstrated in the companion paper for artificial data sets and solar wind models. Here we investigate three relatively well-observed type II events with estimates of shock parameters from LASCO/SOHO observations of coronal mass ejections (CMEs) or other data. Using these parameters we obtain reasonable qualitative and semiquantitative agreement (25-40% correlations) between the theory and observed dynamic spectra. Then, using an iterative downhill simplex method with two assessment parameters, we extract model shock parameters that increase the agreement between theory and observation in terms of relative flux levels, spectral intensifications and drift rates. The extracted parameters agree qualitatively and semiquantitatively with the parameters (speed, size and expansion index) estimated from CME observations for one of the studied events. The extracted parameters agree qualitatively with the remaining two events and yield new initial shock speeds. The agreement between this multiprocess theory and observations is promising for these first quantitative comparisons performed here. Quantitatively, the bulk of the radio emission agrees to within 5 to 10 dB with observations, with the theory typically overpredicting the intensity of bright spots in the dynamic spectra. The methods and analyses presented here show potential for the remote inference of CME-driven shock parameters and the prediction of radio and space weather events.