Flavors like Modoki are actually represented by a continuum that also scales by intensity.  In other words, looking a the strength of Nino-3.4/ONI is a good proxy for the flavor (and vice versa).  However, the SST anomaly patterns for El Nino events tend to be more distinctive than for La Nina events which look more similar to each other.  It is especially not clear that La Nina events should be classified by flavor.  In my opinion, it’s more instructive to just examine the intensity of ENSO events as measured by a variety of indices.  This also makes better sense from an prediction perspective because it’s often the strength of the signal that scales with impacts (signal-to-noise is a measure of predictability).