It has nothing to do with using a brain. The focus confirmation light is a poor aproximation when the dof is thin. There is a range where it light up and that is assuming the "auto" part of the focus is actually in range and correct.

A good coarse screen with a split prism is far more reliable ...and the way to go.

Btw ... on both my F100 and D700 the Zeiss 85 is unfocusable using the focus confirmation light. It stays on for a good 1/4 inch turn of the lens ... which is, at minimum focus the difference between the eyes or back of the head in focus.
And the screens of the F100 is so clear and bright that there is on way on earth to engage "the brain" and focus on the ground glass. Pathetic situation .... fixed by the aquisition of a F3 with a good split prism screen. Now life is good (and sharp) again!