the rear mount end cap cracking i've seen before too.
when i worked for yale and caterpillar we saw it more then we would have liked, and the reasons for the cracking varied anywhere from just bad welds to stress cracks from mast racking where the cylinders were out of adjustment to operator abuse from running into stationary objects.

It's hard to put a definitive reason on why it happens sometimes but there are things you can look for when inspecting the causes for the damage.
- check the mast and make sure both cylinders travel evenly and bottom out at the same time when fully extended and retracted also. (the mast racking i spoke of and was mentioned by kevin)
- check the forks and carriage for damage from impacts, forks may be curled up on the ends or the carriage bent even in the slightest.
- if you can, watch the operators while they are loading up product in the racks or areas where they work, see if you can find a cause in that area. I have on occasion caught drivers doing the dasterdly deed while i was there at the customer site before.
- check the hydraulic tilt pressure and make sure it is at correct relief, over pressure can also make a cylinder crack over time.

it could also be a combination of all of the above or even something i may have not mentioned here.

As for repairing them? Back in the day when safety was not as clear cut as it is today welding them back up was almost always done without question but in todays material handling world safety is now one of the more prominent factors when working on equipment and welding on 'load bearing' structural components is not the first choice anymore. It generally requires factory approval and should be done by a certified welder. Weigh those factors into the repair and in alot of cases just replacing the cylinder is more economical and cost effective and safer than just welding it back up and getting it certified.

It would pretty much be your decision and your company's and ultimately the customer on wether or not to follow through with the repair v/s just replacement.

But with all that said, most of the time in my experiences we almost always found it to be because of bad weld penetration causing a weak spot in the weld and most of the time the masts were out of adjustment causing a bad racking effect. (something the pm tech should of caught during pm's and corrected)