The argument, I think, is more nearly that they should have either slowed down the phone from day one (never promised more than it could be certain to deliver after a couple years) or made the phone bigger (include enough battery to maintain the promised performance indefinitely). In engineering, the goal is often to observe failure as close as possible to the computed target, as that shows you are minimizing waste and designing optimally. But doing the worst possible design that meets the goals, while generally understood the engineering goal, may still be counter to the expectations and understanding of the consumer. But I'm also a fan of my iDevices.