Certainly true -- though the commonality in underlying physics lend a certain grounding back to a common reality. The preferred techniques will influence which shapes and which movements are favored, but speed and straight movements are still favored in most attacks; redirection and circular movements favored for deflections (though not all defenses).

That's true--but some arts emphasize attacks more than others. Aikido, for example, is very fond of techniques that combine attack and defense.

How much, then, does the real effectiveness of different styles of unarmed techniques come down to the fact that people's bodies are different, and different techniques are better for different people?

That's absolutely true, and different instructors deal with this reality better than others. For every technique in a particular style that a particular student can't perform, there will be a substitute that has a similar effect but works better for the particular student.

Our MMA instructors are very aware of this, because all of MMA is performed at least at dancing range (and a lot of techniques involve being pressed against your partner), but they vary in how well they communicate alternatives.

And you are certainly right that sword techniques can be intensely circular. esrblog and I study a form of historical fencing (supposedly 15th c. Italian). The shield motion techniques we are taught are all circular, and sword recovery moves are also. This is based upon body physics, again; the circular recoveries are to redirect the remaining force of a blow to spare the wrist, and the circular shield blocks give greater power with less use of muscle.