Continuing the discussion from Inference hidden components:
As discussed in the linked topic, there is a ‘bug’ with the way the inference engine treats hidden geometry. As @Geo remarks, “the Inference Engine dynamically retains a certain number of previous inference points.
If the geometry is subsequently hidden those previous inferences briefly remain in focus. That is, until user actions shift inference focus to other things.”
However, I would argue that if a user toggles hidden geometry visibility off, she expects all reference to that geometry to go away. Certainly it’s bizarre when a tool snaps to invisible points. The engine could remedy this by flushing stored inferences when visibility is turned off. It works with layers anyway!
Here’s a video of the ‘bug’: https://youtu.be/b7O6Bh86Rs0