I loved watching for the moment the topology of the creature changed from a sphere to a torus. Not exactly sure when it happened, but there was a point where it wrapped itself in half and seemed to fuse.
Generally, in development, it's the lineage of the cells that matters the most. Skin and neuronal cells have a closer lineage than the gut cells do.
Neural tube-like structures are pretty ancient. Hallucigenia, I believe, is the oldest fossil (-0.5 BYA) that shows the differentiation of a single neural tube-like structure [0], right back to the mid-Cambrian Explosion. We are the ancestors, likely, of that animal. It seems that bi-lateral symmetry is a highly advantageous body plan, as it's evolved a few times in separate clades that weren't that before. As to the reason why bilateral symmetry is good, it's a good question! As for the difference in why neural tubes are better, it seems that the ability to centralize the information exchange between the body and (what may pass for a) brain is a better fit with a single trunk. Other animals do have other arrangements, but the one we have has the most diversity of subsequent animals forms. It seems to be a more flexible arrangement and evolution seems to have preferred it. Granted, the neural tube idea has been through a few mass extinction events, so there is a considerable amount of luck involved too.
[0] For reference, the reason that arachnids are different than other 'insects' is because they have dual ganglia structures that run down the rostral(head)-ventral(tail) axis instead of a single neural tube. That divergence took place before Hallucigenia, I believe.