Here's another test that might seem more applicable to the real world - stab something, hard. Doesn't matter what it is, phone book, cutting board, tree, whatever. If the knife closes on you and cuts you, the lock isn't functioning, and the knife isn't safe to stab with. I know I'm not the only one on this forum who's been cut doing this (I stabbed a piece of styrofoam).
Spine whack is a similar test, but, if done properly, you don't get cut if the blade closes.
Sometimes, in real world use, the spine of the blade will be bumped by something. Some knives close when this happens, others don't. I assert that the knives that will not close when this happens have superior (i.e., functioning) locks. Since I value my fingers, I refuse to use a knife that doesn't effectively lock open.
They're your fingers, and if you want to assume that the lock on your knife will protect those fingers, yet are unwilling to test it to actually verify that, then you are assuming a certain risk as well. You might get lucky, many have; you might get cut, many have. If there did not exist a folding knife lock that could consistently maintain its integrity while being spine whacked, we would reasonably conclude that this is an unreasonable test to expect knife locks to pass. But this is not the case. Many folding knife locks can consistently maintain their integrity under a spine whack; these knives are, therefore, superior (in their ability to remain open rather than close on your fingers) to those that can not.
If the deadbolt on your front door fails when you bump into the door, you are free to conclude that this is abuse and not a valid test, or you can replace it with a deadbolt that actually will keep your door from opening that easily. It all depends on what you expect your deadbolt to do, but some are stronger than others.