Fitts' Law measures how accurately people can move from some point to a target using a pointing device. It's a mathematical equation.
Targeting becomes way harder if you can't see the target, given it's a pointing-based targeting mechanism.
Other mechanisms exist!
Audio targeting would allow a person to select an item by making a predetermined noise.
Better yet, make it 3D and include head movements.
This would work particularly well with the audio cues in the audio notes.
Easy way to increase the targeting effectiveness: make the targets bigger!
MacOS dock does this, though it negates the effect by also moving the target, making it harder.