Nothing drastic I'd say. It's a continuous stream of small improvements just accumulating with each release, and someone just noticed a few releases away from a previous publicized-bad-capabilities release that there's major improvement between those points. So it looks like something major only due to the spacing between the capability surveys on the release timeline.

It was drastic and immediate. It switched with the latest versions of opus and codex. It's why openclaw is popping off. The models became usable.