I don't see how the paper you're linking to support Hinton's supposition. It's a study of topology, it doesn't aim to show that biological neural networks learn

I don't see how the paper you're linking to support Hinton's supposition. It's a study of topology, it doesn't aim to show that biological neural networks learn via backpropagation.