For a definition of intelligence, a multi-dimensional perspective is needed. Read this: https://medium.com/intuitionmachine/deep-learning-system-zero-intuition-and-rationality-c07bd134dbfb
This should give you as sense that progress needs to be made in multiple dimensions. AGI concerns itself mostly about the autonomous dimension (that DeepMind in a recent paper explains in further detail).
I think in your response you argue for an ‘autonomous’ definition of intelligence. That is just one kind of intelligence and the community is working on several kinds but they don’t actually articulate the difference and just label it under one word “intelligence”. So, it’s really a matter of education in that most people actually haven’t really taken the time to understand the word.