I want to highlight one sentence from the paper:
The materials involved in domain specific savant abilities often involve human codes such as arithmetical structures, written codes, calendars, music scales, 3-D regularities, and natural taxonomies that all feature structural redundancy.That's an accurate statement of present fact, but it also raises a fascinating historical question: Were there any savants/autistics in existence say ten thousand years ago, and if so, what were those savants/autistics perceptually attuned to? It sure as hell wasn't arithmetical structures, written codes, calendars, music scales, 3-D regularities, or natural taxonomies.
That might seem like a tangent observation at first, but it actually goes straight to the heart of the matter and should make one think. It should especially make one think when one is trying to justify a bunch of neuroscience gobbledygook.