The thing about liner locks and framelocks is that the perceived simplicity -- basically just a slit in the liner/frame plus an angled tang -- hides enormous sensitivity to all kinds of factors. Most importantly, the sensitivity to geometry, where every blade torque and handle flex provides a new opportunity for slippage. Or, mechanical simplicity leads to failure mode complexity, inthis case.
By contrast, the "complexity" of milling some pockets, adding a slot in the handles, and putting in some springs seems to have freed the axis lock of the sensitivity to those myriad failure modes. Adding a tang shelf and liner projection -- more complexity -- leads to the compression lock, which also adds complexity to gain reliability.
I believe people are way focusing on the wrong thing when they focus on just marginal additional lock complexity, instead of practical reliability. Simplicity is not a goal for simplicity's sake. Simplicity usually helps reliability. When it doesn't, it's time to add in features to help the reliability -- like the axis lock, or compression lock. The axis lock, for example, shows every sign of being as reliable as possible with today's technology, with the proviso that you may have to replace a spring down the line.
I'll also add that, although a recent thread shows that there is no consensus on this, I believe the framelock has such reliability advantages over the liner lock that it should be mentioned among the other reliable formats, specifically because of hand reinforcement. I haven't quite squared that away with the fact that the framelock can introduce a big failure mode to counterclockwise torque if not carefully designed. But I carry several framelocks with confidence.
Joe