IMHO, I think for reliability : strength : usability : price in a high-volume production folder, the AXIS-lock has most other locks beat.
- It's completely ambidextrous, along with easy one handed opening, one handed closing
- Fairly simple requiring less of the proper fit of a linerlock (therefore more likely to get a functional piece OOB)
- Much less susceptible to failure under dirty conditions than a linerlock or lockback
- Has redundant springs
- Is inherently strong (probably even stronger now with the introduction of an 'integral' AXIS lock, although in use, the blade or blade stop will give way first in either rendition)
- Not susceptible to torque, or other forms of unintentional disengagement (even with fingers positioned on both sides of the blade, it's rather hard to disengage intentionally through use, and I doubt that anyone will stab anything hard enough deep enough/viscous enough to first penetrate the length of the blade, then penetrate past the front of the handle, then move their hand in such a fashion as to allow the material to congeal around the lock area)
- Through use, the lock engages more securely (locking bar rides up the ramp) (also, lock seems more wear resistant than a linerlock over the long run, tho' I don't have evidence to quantify this)
Cons include being harder to clean, and only one company produces them limiting choice.
You can probably tell I'm a fan

, but I think that it's hard to beat in a working folder being produced in high-volume with all the attendant QC problems that are bound to occur.
That said, I think lockbacks and linerlocks are fine locking forms when executed properly. However, seems to me that the AXIS locks are executed more consistently (esp. coming only from a single company), and have surpassed other locking formats in usability. But I haven't had experience with compression locks, and balisongs will surpass AXIS locks in reliability.
This is looking at production folders; I think custom folders will inherently be more reliable (not sure though).