Here's what has been proven regarding the P320's striker safety...
- The Striker Safety needs to move only 0.036" to disengage from the striker (brand new P320).
- That correlates to a trigger pull of only 0.09" to fully disengage striker safety (brand new P320).
- The trigger breaks and the sear releases the striker at 0.201" (brand new P320).
- That means the additional trigger pull past striker safety disengagement needed to have the sear release the striker is 0.11" (brand new P320).
That means it's certainly possible that if the the trigger is barely moved 0.09" or perhaps even less if striker safety and/or the striker is broken in/worn, and/or the slide to FCU/Grip fit is loose, and then if the slide is torqued vs the FCU/Grip in a holster like it appears is happening with many of the LEO UD's, that ultimately the striker could be released ("wiggle fire").
The P320's striker safety is rather unique design, and nowhere near as robust as the much more commonly used vertical plunger designs (P365, Kahr, Glock, CZ and many others etc), which also includes the fact that the FBI Ballistics Lab reported the striker safety's torsion spring is only captured on one leg which means means it's feasible the P320's striker safety spring could fail to control the striker safety without breaking if the non-captured spring leg were to move and there by lose tension). The P320's optional manual safety is also inferior in that it only blocks the trigger bar, and not the sear.
I'm not a Sig hater seeing I love my P365 with Spectre Comp Slide which has a relatively tight slide to FCU/grip fit and it will simply not "wiggle fire" at any trigger pull prior to the break.
So, IMO the P320's striker safety timing is possibly too aggressive, and when combined with MIM parts tolerance stacking and a possible lack of effective MIM parts QA in some cases, and that's in addition to some sloppy slide to FCU/grip tolerances, then there's currently a cause for concern in an admittedly small percentage of the many P320's/M18's/M17's produced (while of course noting that some percentage of the UD's were operator or armorer error etc (10mm parts in 9mm guns etc).