On "science that is greater than the sum of its parts", the post hits on key ingredients:

+ understanding of problems from multiple perspectives and across multiple dimensions;

+ pragmatic engineering for high-quality execution;

+ touching the real world, doing it at scale;

+ demonstration and iteration, always;

that can together enable "execution-driven science." In practice, much of that R&D progress is posted at https://pasteur4d.ai/research and other offshoots of https://arxiv.org/pdf/2112.03235