The first applications appear really simple because they illustrate cognitive abilities that imitate planning/model-based RL related to neuroscience/psychology.

The first applications appear really simple because they illustrate cognitive abilities that imitate planning/model-based RL related to neuroscience/psychology. It doesn't mean that meta-RL won't scale up with more computation (see http://www.incompleteideas.net/IncIdeas/BitterLesson.html http://www.incompleteideas.net/IncIdeas/BitterLesson.html).