The formula is not perfect for sure and there are many variables to throw in the mix. For instance, the single biggest destroyer of throats is over-heating. A 30 BR fired like a belt fed machine gun is not going to last 8000 rounds. Also, the formula does not take into account they variable in terms of powder charges and bullet weights. Lastly, the number it produces is an index, not an absolute figure.
The 6BR is indeed good for about 4000 rounds IF IT IS FIRED SLOWLY AND NOT OVER-HEATED. the majority however, are not. 2000 - 2500 is a typical accurate life.
The 50 BMG has an index in excess if 1400, yet it has a reasonable barrel life, likely because few can tolerate or afford to lob a bet of ammo down range in one sitting. I think this formula has flaws with large diameter bores to be honest. A "bore factor" needs to be added me thinks.
Also, this formula represents "functional accuracy". I am convinced that the most accurate life of ANY barrel is in its first 200-500 rounds. After that, a degredation in accuracy occurs that may be very slight in some, or substantial in others, but it does happen. The formula really represents what i would arguable call "F-Class Accuracy", that is, sub . 5 moa accuracy (I'm using a custom barrel as a term of reference, as many factory barrels are incapable of this accuracy even when new) . Many old 6.5X55 rifles have thousands of rounds down them and are still more than accurate enough to hunt with. Accuracy and barrel life are relative to purpose and expectation.
Kombayotch: I agree with everything except bullet coatings. There is a distinct lack of objective credible research out there to support that. Also, the benefits are often offset by the need for more powder to maintain an equal velocity.