An actual failure in use was what got me looking into liner lock reliability. I don't remember how long ago it was, maybe 7 years ago, and I had been hearing for about a year the first rumblings of liner locks failures. And one of my favorite posters on rec.knives had, little-by-little over the past year, decided liner locks were unreliable. I hadn't thought much about it until my first failure.
That first failure came on a custom knife from an extremely good maker. I had already noticed that when I held the knife firmly, the liner lock disengaged ever so slightly. One day, I was holding the knife firmly and the knife disengaged -- luckily, I was already suspicious of the lockup, so as soon as the lock released, I backed off and didn't get cut.
At that point I re-considered the position of my friend from rec.knives, decided he knew what he was talking about after all, and started paying closer attention to liner locks. I've never had a liner lock failure in use since then, which I attribute to the fact that I test my liner locks so thoroughly (and the fact that I stopped carrying liner locks at all 1.5 years or so ago
). For a while there I was the liner-lock-failure posterboy, so people would send me emails about their particular failures.
I understand that the spine-whack test is not universally accepted, and I understand why, although I feel it's a good test. However, I strongly feel that torquing and white-knuckling are very important tests, unless you only use your knives for very light use, or as a gents' knife. I mean, haven't you ever cut, say, stiff cardboard, gotten the blade stuck, and torqued it out a bit?