Maybe they could "remember" where your eyes were looking at a few milliseconds ago and click on that, but then they wouldn't know the difference between that or maybe you actually did mean to click in the center of the content area.

The problem is the Vision OS has no cursor. Cursor position is critical. Your eyes move around a lot more than your cursor moves, but in Vision OS, cursor position and eye position are the same. Very annoying. Maybe they could put a little cursor in the center of the screen that you could move around with another gesture.

If you want to stick with the old mouse paradigm. I rather think people are going to get used to looking at what they want to click on for slightly longer very quickly.