Sutton himself still thinks the current methods are insufficient. The question is not whether the Bitter Lesson is true, but whether we have algorithms that scale sufficiently with compute to warrant no further research.