It would be very interesting to see a detailed analysis of how historical ONI values and El Nino/La Nina episodes compare when using SST anomalies computed from ERSSTv4 versus v3b. Other scientists at NOAA also seem to study SST using the "cold-biased" satellite-based data explicitly excluded from ERSST (see http://coralreefwatch.noaa.gov/satellite/sst.php for example). What do ONI values look like when computed from this data?