The points are roughly distributed into 1/3 El Nino, 1/3 neutral, and 1/3 La Nina mostly because the threshold (+/-0.5 degree Nino3.4 index) is close to the boundaries delimiting the upper, middle, and lower thirds of the available data. Your interpretation of the data is correct, but you make an important point when you mention the limited data. It's too limited to make a very confident statement, but does give us an idea of the relationship between the winds and the ONI value. We have not repeated the study with pre-1980 data.