That's one of MANY known issues with data mashups.
What is occurring at an incomprehensive scale is the complete data capture, correlation, classification, and semantic analysis from every major point of communication.
The technology when I entered the industry - a 20 megabyte hard drive cost $2,000 in 1984 ($64K now). A 330 gigabyte mainframe disk pack cost upward of $60,000 in 1988.
Today, we have multi-terribyte in-memory, massively parallel, multidimensional data access platforms in the private sector. Just think of what can be done with unlimited time and resources (>$80 billion/yr) of the the federal guberment!
Duh, a 330 megabyte diskpack cost $60K. A gigabyte of storage in any device didn't exist in '88.
Now, hot-swapable terabyte RAID units for PCs exist for a few hundred dollars.