I firmly believe that using proof loads repeatedly will cause very early failure.
I agree. Testing anything to limits well above its intended design range, REPEATEDLY will result in eventual failure. And that failure will happen before (by count number of cycles, or shots) it happens operating within design parameters. That is "early".
The gun is meant to stand a proof load (about 33% overload) ONCE.
That is the standard we use today. A sensible one, considering that if you test a product to destruction, you no longer have a product you can sell. Pass proof testing (once) and its good to go. Keep proof testing and eventually it will fail. Might be the 3rd, or the 33rd or maybe even the 333rd time, all we know is, at some point it will fail due to repeated overload.
if you can accurately predict when the gun will fail (shot #) with proof loads, then please tell me next week's winning lottery numbers. I'd share the jackpot!
The concept of proof testing" comes down to us from the days of armored knights. Look at the (surviving) sets of "high quality" armor plate. You will often find a "dent" in the breastplate. Deliberately made by the smith, who shot the armor (with a bow/crossbow originally and later a gun) to "prove" the armor would withstand the impact.
Wealthy noblemen, kings and princes weren't going to lay out serious coin for armor and just take the word of the smith on how good it was. They wanted proof! Eventually the powder charge (and ball) used to test the armor became known as the "proof load".
The same concept in gun making is the load the gun is tested with as proof it (the gun) will survive.
It is ridiculously easy to create handloads that hugely exceed industry standard proof load pressures. This is the main reason it is so vital to pay attention to what you are doing when making ammunition.
) remains one of my most accurate rifles to this day.