How do they close the feedback loop?

In all of these systems I've seen so far, there is a ton of calibration needed.

The user should be able to do that.

Edit: say it is a two way street.

System takes input and renders it right back as a shock, or vibration, something.

There then is an exchange.

System presents its message, user present theirs.

Simple things, like shock, or pressure, vibration on and off.

Then levels.

Then modes.

Together, that results in "thought as action", much like we get when we do things with our bodies.

Once there, mapping that to things like cursors, could be standardized, IMHO.