Only one of those pictures is clear enough to really tell anything...but my nerdy bored on a Saturday morning over engineered 2 cents...
Assumptions:
1) There is an obvious stress riser where the small stem diameter meets the larger boss/spring locator. A larger fillet radius here would help alleviate that stress riser.
2) The PV is 6061-T6 aluminum. Why? It's common, readily available, versatile, lightweight, easy to machine, and CHEAP...also...most jet ski folks aren't in the market of making parts from exotic alloys. 95% of your "billet" aluminum parts in any aftermarket are manufactured from 6061-T6.
This looks like pretty textbook fatigue failure. You can very clearly see the point of initiation, crack propagation, and finally the fast fatigue fracture. (meaning it starts to crack, the crack spreads, and it finally breaks suddenly). I suspect it was accelerated by 1 of 3 things, or a combination of the 3.
1) Over torquing of the PV cap bolt damaging the stem at the stress riser.
2) Excessive clearance between the PV and the cylinder causing it to rock (and fatigue).
3) Piston contact of the guillotine, although this likely happened after the stem failure.
Why do I say excessive clearance? Because if the clearance was too tight causing galling and stiction, the PV would stay stationary, or stick. Then, the only forces acting on the stem would be the spring force, and the force created through exhaust pressure filling the bellow. Both of these are very small forces that in no way could cause tensile failure. Now, if it were excessive clearance causing rocking; this *could* cause small shock loads in the PV and over time accelerate a fatigue failure at an already weak point (the stress riser).
As to water causing the failure due to hardening or making the material brittle? Not likely. Actually...I'd say no way, no how. First, 6061 is not very quench sensitive compared to say... a high alloy steel (where brittle failures are more likely in heat treated specimens). Second, in order to heat treat and quench and temper aluminum alloys, you need to get the material to ~900*F or higher. These PV's are held within the cylinder, which is a giant heat sink. So while they are exposed to small amounts of hot exhaust gas, I'd feel safe to assume their temperature is equivalent to that of the cylinder under operation...ie... nowhere near the 900*F it takes to heat treat aluminum. If it were that hot, the rubber PV bellows which are filled with exhaust gas would melt very quickly.
My recommendation? Call Dasa (I know, "call your engine builder" is becoming the most popular response around here) and get the proper clearances and tolerances of the PV to cylinder, the PV to piston when installed, and while you're at it, ask about the acceptable piston to bore clearance. Measure your cylinder and your new PV's to make sure you're within the specified range. Measure your bore and pistons, you may need to bore the cylinder after the rings made contact with the PV's. Once all that is OK, make sure you have proper clearance between the new PV's and the pistons when installed. Finally, when installing the PV caps, just snug the bolts with a drop of red loctite.
Good luck and let us know what you find!!