I have just read Alice Gordenker's July 16 column — "So, what the heck is that?" — about ishigaki construction at Japan's castles. The article conveyed interesting information about engineering design, but I think that the castle expert quoted in the article, Yasuhiro Nishigaya, has been misled about the reason for the round or curved walls of European fortresses.
The round walls were an answer to the potential damage from early artillery sieges, mainly from catapults. The easy way to open a breach in a wall or tower was to knock off an edge. Rounded edges made the corner towers less likely to tumble in a siege. I don't know whether that also led to ill-aimed shots glancing off the targets.
When gunpowder artillery came around in the late Middle Ages, high superstructures that until then had given an advantage to defendants increased the risks, so the walls were lowered. Castle residences with right-angle corners, rather than fortresses, were built.
In the 17th century, the new look of fortresses was the Vauban type, which gave priority to firing ability — "covering" all parts of the fortress from at least another point within the fortifications. Vauban also had a lower superstructure, reducing the size of the target for attackers trying to bring it down with cannon. If I recall right, the walls up from the moats are inclined. In Europe, fortress construction did not incorporate any recognition of the possibility of earthquakes as such before the 18th century.
The Vauban fortress might have served a function similar to that of the Japanese castle wall, but ishigaki is definitely more beautiful to look at than an inclined earthen embankment topped with a breast-high crest of masonry.