Most climate models are coupled to an ocean model, so I'd be interested in whether these dynamics reproduce themselves in the models that the ENSO forecast team already considers.  Just so we're clear, what index are you looking at to define the onset of discharge?  (region, dataset, etc)  

We did see evidence of discharging near the peak of the El Nino in 2015-16... something we wrote about here... but this may not be precisely what you're referring to because I see in your paper you're referring to lead times up to 18 months.  

https://www.climate.gov/news-features/blogs/enso/life-and-death-el-niño

Ultimately, we had a weak (barely there) La Nina during 2016-17 via our ONI index.  

Thanks for your input!