Researchers have found it is not economic influences that limit the growth of digitally stored information on the Internet, but the human brain. Reported by Springer is the study (pdf) which considered some 633 million files, constituting 675 TB of data. This represents every file there is an outgoing link to from Wikipedia and dmoz. The file types include applications, text, images, audio and video.
The researchers looked at specific characteristics of the files, such as the bit rate, resolution, and length. These data were then plotted by how often they occur for each data type. The graph shows a decline as the axes increase, but is without an exponential tail at the end. If economic factors, such as the cost of hard drives, limited how large or how high of quality a file was, there would be such a tail. Further examination of the graph showed patterns which match the Weber-Fechner law. Basically what this law says is the noticeability of a change is follows a logarithmic curve. For example, increasing the resolution of a low resolution image is more noticeable than increasing the resolution of an already high resolution image.
What this implies is the information on the Internet cannot grow faster than what our brains can handle.