A force layout has no idea what a good drawing looks like. It only knows three local rules: nodes repel each other, linked nodes are pulled together by a spring, and something weak drags everything toward the centre so the whole thing does not drift off screen.
Run those to equilibrium and structure falls out for free — clusters become visibly clustered, bridges become visibly thin. Nobody designed that; it is just what minimising the energy looks like.
The honest limitation: the result is not stable or reproducible. Same graph, same code, different starting jitter, different picture. Which means a force layout is a good tool for exploring a graph and a bad one for referring to it.
The layout of the graph on this site is exactly this, with the node colours coming from oklch. On why the notes are small enough for it to work: zettelkasten.