The formula offered by the devs showed full fatigue being reached in 2/3 the time that it initially takes to enter into fatigue. This test had gone on for six hours before I even started this portion of things, so I was burnt out and ended up mining slower during this portion of the test, evidenced by my numbers. However I did the math after I called it quits, and it turns out that I actually mined at 2/3 my original speed while in fatigue. At this rate, I still should have hit full fatigue within four hours if the 2/3 ratio in the original formula still applied. This is not what happened.
Full fatigue... is a lie?
After a total of ten straight hours spent mining, I called it quits at this point.