Hi Paul,

There indeed was a noted change in the Pacific after 1998, often considered a change to the negative phase of the Pacific Decadal Oscillation (PDO) or Interdecadal Pacific Oscillation (IPO), which is associated with more frequent occurrences of La Nina relative to El Nino. However, the degree to which we can call this a "cycle" is debatable, as a large component of this variability may be driven by chaotic ENSO variability. In any case, this gets at the "internal variability" argument, and so if internal variability really is the culprit, then yes, we should expect a reversal that coincides with more frequent El Ninos. However, we still would need to address why this recent period has been so unusual because climate models should, in principle, capture this internal variability. (Not enough decadal variability in climate models? Or was nature just particularly extreme in recent decades?). 

Regarding the MJO and Kelvin waves, yes, these influence the Walker circulation, but I'm not aware of any way that they should have an appreciable effect on trends over the course of decades.