What's subjective to you is pretty objective to me - the health impact of unhappiness, for example, will reflect in healthcare costs. Which btw makes up 15% of US GDP. Now is that even real GDP or just a manufactured problem - with doctors and lawers and insurance guys all making money off of health problems of consumers.
See this - http://en.wikipedia.org/wiki/File:International_Comparison_-...
For the same baseline level of internal well-being, people will respond to the question "how happy are you?" differently depending on their culture. Do you understand how this presents a problem when measuring happiness across cultures? That is why I refer to it as subjective. Cultures change over time, so a change in the HPI might simply be a reflection of that.
That being said, I do not believe that just because a measure is subjective, it is useless. Of course it's useful. It's just important to understand what exactly you're measuring.
So taking into account cultural variations is meaningless when what I'm interested in measuring is the end result.