I'd like to come back to this as I believe it to be rather significant to this discussion.
How many of you that have suffered engine failures have actual data logs listing the ECU's input sensors readings leading right up to the point of engine failure? How have you come to the conclusion that the engine failed due to 'too much boost'? Are you just guessing?
There are so many factors that are relied upon to accurately calculate ignition timing, fuel injector timing and fuel injector duty cycle, if the calculations are wrong particularly during a high load and high engine speed condition, for whatever reason, severe detonation will occur and something is likely to fail and that something is likely to be the headgasket, a piston or a conrod. IMO a bent conrod is more likely to result from the increased cylinder pressure and the hammer like shockwaves caused by severe detonation than it will from increasing the intake charge alone. Detonation is more likely to be the engine killer when increasing turbo boost pressure.
Severe detonation could occur for a number of different reasons i.e. A low battery voltage condition that reduces fuel injector operating speed resulting in insufficient fuel being delivered, the ECU not receiving the WOT position signal from the TPS or even the knock sensor failing to report knock that is actually present resulting in the ECU over advancing the spark.
Take the MAP sensor for example, a wiring fault causing an electrical resistance or air leak causing a difference between the air pressure in the intake manifold plenum and the air pressure seen by the MAP sensor, would most likely result in the ECU receiving an incorrect feedback voltage from the MAP sensor. If the ECU only receives a feedback voltage of 2.5v instead of the intended 5v, then the ECU would deliver a totally incorrect amount of fuel from totally the wrong area of its fuel map. The ECU would also incorrectly calculate engine load and so would provide a totally incorrect amount of spark advance from totally the wrong area of its spark map. The engine would then suffer a situation of insufficient fuel and too much spark advance right when the engine needs more fuel and an earlier spark to be delivered to avoid detonation and stay alive. All this could simply be caused by a small split in a hose or a slightly corroded electrical terminal and probably wouldn’t even flag up a fault code.
Without firm evidence to the cause of an engine failure, I think it's a bit naive to blame it solely on the amount of boost pressure being run. I'd be more inclined to blame it on the calculations the ECU had made (whether due to an error in a multiplier map or incorrect map lookup based on an incorrect input sensor reading) to compensate for that given amount of boost pressure and engine situation leading up to and at the exact moment of engine failure.
I totally agree and I personally would spend my money on some decent well respected engine management before I even though about replacing pistons and rods. Although it needs to be remembered that 100's of hours would have been spent developing the maps for a stock ECU, so no matter how good your tuner is unless they have been afforded a similar amount of time to map your ECU, reliability may always be an issue.
As hardcoreep once said: