I think the worldview problem with measuring inflation in terms of CPI (consumer price index) is that it presupposes that inflation is some kind of natural, inevitable aspect of any "normal" economy. If anything, money which is denominated in terms of some sufficiently costly to produce commodity, such as precious metals, should on balance be deflationary. If a quarter piece of gold could buy a week's worth of groceries in 1900, the same amount of gold should be able to purchase much more groceries today because, despite the modest increase in the world's gold supply in the last 100 years, the technological innovations which drive down the cost of producing food have followed an exponential curve that makes the world's gold supply look flat by comparison.
I expect that an empirical study could be performed to vindicate this view by estimating the amount of consumer goods which could be purchased with a unit of gold in 1900 versus the same unit of gold today. And this neglects the increase in the number of available purchasing options facing consumers that enables them to purchase what they want, not just the quantities they want.
I have been thinking about why the government (and the central banking system) can engage in inflationary policies without people minding and I don't think it's just that inflation is difficult to understand, because it's not really that hard to understand. I suspect that a more important reason is that it is human nature to form our future expectations on the basis of our past experiences. In the presence of exponential economic growth, this means that people's expectations of their future wealth will, on balance, be too low. If you grew up in the late nineteenth century, automobile transportation - which we take for granted to the extent that we treat it as a basic necessity (just look at all the hand-wringing about oil prices) - would have probably seemed an exotic luxury and the development of jet transportation was simply inconceivable and unforeseeable.
As remarkable as our past growth has already been, the trends indicate that we should continue to expect accelerating exponential economic growth for the foreseeable future. But most people, when polled, have a gloomy outlook about their economic future, expecting to be worse off in the near future than they are today. The reality is that as the world becomes more connected, trading in the world market will result in a massive decrease in consumer product costs (making us all wealthier) even if there are no future technological innovations. But we also have good reason to expect future technological innovations, the most spectacular of which will probably come from the fields of medicine and biology.
This means that, while the opportunity cost of hiding money under your mattress is always non-zero (it could have been loaned out and earned interest on capital investments) this should not result in a loss in real purchasing power unless that currency is being diluted by expansion of its supply. In other words, if you hid $1000 of gold under your mattress in 1900, it would have missed out on the growth opportunities of capital investment (at a modest 4% growth rate, it could be worth $50,000 - in 1900 dollars - today), but that gold would still be able to purchase many times more consumer goods today than it could have in 1900 because consumer goods have become so much less expensive. If, on the other hand, you hid $1000 USD under your mattress, you could purchase far fewer consumer goods with it today than you could have in 1900. Unfortunately, I think we don't appropriately distinguish between opportunity cost (which causes uninvested money to become less valuable relative to invested money over time) and inflation (which causes money to become less absolutely valuable over time).