(think proof testing of a firearm where a firearm is deliberately subjected to overpressure (abusive, yes))
Ugh, they actually measure pressure when they do those tests - doesn't do much good if you don't even know how much gunpowder you used in the shell or the pressure that resulted in the failure. And it also wouldn't tell you much if you used the same brass case or the same chamber damaged by a previous shot for every test (cumulative damage factor).
To compare the videos to real destructive testing, which is by the way being used less and less all the time, reveals a poor understanding of real testing. Google "boeing 787 wing box" to see a real destructive test. This is a test that can tell you something. Even the simplest tool test is done with much thought & effort to measure the quality that they want to test for. And they can defend their results. Feel free to defend the videos if you wish, but there is zero real product or material tests done in a similar fashion anywhere in industry or by testing agencies. Not one.
If you want to test how long a knife can be beat on with a steel hammer using varying impact locations and varying impact forces, then you should test a number of specimens before making conclusions, due to the random nature of the test.
If you want to test how well a knife can be used to batoned by a log into wood, then you should test by batoning with wood into wood.
If you really want to get fancy and learn even more, than devise a way to make the test repeatable (this IS an actual requirement in real testing, you know).
One test will not tell you the same thing as the other. It is likely that some could withstand 10 years of being hit by wood, and break after 10 hits with a steel hammer into a concrete block.
And it is entirely possible that the knife that breaks when used as a pry bar is the strongest pry bar, but you won't know that without measuring the force. You might think the knife that bent and then slipped out of the pry is better, but it failed (bent) with much less force than the knife that broke. And would you rather have a broken tip or a bent tip on your "survival hard use" knife?
I'm not trying to set an unrealistic level for testing, but there should be some thought on what you want to find out, and how to make a test measure it. BEFORE testing. And without repeatability, a test really doesn't tell you very much.
