My point was this: If you can be that accurate with a gb estimate, then there is your aim, no need for any measurements or estimated lines trying to pinpoint a value in the middle of the table, like that "1.7" value shown in the diagram.
Also, being off by a couple of millimeters on the gb...