The primary issue I have seen with your work is that you tend to get lost in the trees and lose sight of the forest, so to speak.......the data is overwhelming as it is for me going back to 1950, nonethless 1870 or whatever the hell you do. There is a paradox that exisits in that on the one hand, we have such a grave dearth of data in terms of sample size, but on the other hand, there is data saturation with respect to the sample that we do work with. Its important to set boundaries so you don't endure "paralysis by analysis" and spend more time sorting data than you actually do forecasting. I have noticed that you tend to exhaust yourself with ENSO composites, which is something that I have moved away from because if anything, ENSO is growing less preavlent as a seasonal driver. ENSO is the foundation of the seasonal house, metaphircally speaking....we have a finite amount of time to complete this seasonal work and if you spend 2 months polishing the foundation, you're going to build an inferior house.