Bayesian model averaging: ranking of model weights and LOO don't match

Yes, the paper (and others by the same group) outright say that. I don’t know why pseudo-BMA is preferred by ArviZ.