Everyone is noting some great thoughts here. And, it looks like someone is putting his thermodynamics to work. But, it is best to address the issue in the simplest terms possible, so those who are not engineers do not pass out.
Mass flow dictates HP. All of the reasons stated speak to sampling mass flow. The only thing left to say is turbos create restriction to exhaust flow, requiring more boost and increasing piston temp. Turbine sizing is a critical component to evaluate power output. There is not a all size fits all rule of thumb for HP loss with elevation since changing back pressure (turbine sizes) greatly influences power output. In the end, you have to do the math, data log, or dyno it. Unfortunately, too many people are focused on compressor wheel flow rates, which is why companies market HP ranges for turbos on compressor wheel size. In reality, the turbine wheel and area radius dictate power level. Only then should the compressor wheel be matched to the mass flow rate demanded by the motor.
In short, think about this in rough terms. If I lose 1 psi MAP, I need roughly 1 psi to get back where I started. Boosted engines, like NA engines, lose power with elevation if boost is static. Electronic Boost Control just ensures MAP remains constant; the gauge only changes. Finally, who cares what the HP is if it feels good to you. Most companies inflate dyno numbers anyway by manipulating dyno inputs and reporting peak numbers.