Comparing loo for two complex multivariate models with large data

Thanks for your response @paul.buerkner. I think I’ve come full circle back to my original post now where I am experiencing a bug in the loo function that doesn’t seem to have a solution.