Not likely being UW has any effect on this. The fact that the piston was replaced actually makes it more suspect, it's not the first time a good part has been replaced with a bad one. If you have a IP gauge remove the second stages, plug the holes and SLOWLY crack open the valve carefully watching the IP, my guess is it will go way high....just be sure to cut off the valve before it gets too high. It would be easy for a less than careful tech to miss this and assume since the the IP "stopped" it was good when in reality, the second stage was leaking enough to stop the increase any farther.
One more test you can do UW is to detune one stage a good bit, breath off of it and watch the other one. If it's an IP leak, the unused (and lighter tuned) second stage will start to dribble air in a few seconds after you take a breath. Take another breath and watch the bubbles from the other second. they will stop for a little while, depending on how much leak you have, it may take a few to a number of seconds before it starts to dribble air again. If it starts again, dead give away the first is leaking.