[Barrin92]

>How is "Baumol effect" different from supply and demand for any limited resource?

It isn't, but nominal growth as a consequence of demand without increase in real output makes you no better off, that's the "fake" part. More practical example than graphics cards is houses. Large chunk of the US market, constantly goes up in price, but not because housing is becoming more productive. It's because money from other sectors spills over. Good for landlords, bad way to measure real economic activity. You'd be better off if you could roll houses from the conveyor belt and collapse prices.

Rising education prices don't reflect greater quality in education, faster teachers, or more graduates, i.e. output, which is what we ought to care about, but just higher spending funded by real gains in other sectors.