Same conclusion. It's too easy to feel that precarity from the early computing age (not enough storage! not enough cycles! data versions of things are special!) if you were there. I think there's some of that going on here on Slashdot a lot of the time.
People in love with old Unix boxen or supercomputer hardware. People that maintain their own libraries of video, but all that's stored there is mass-market entertainment. And so on. It's like newspaper hoarding.
Storage and computation are now exceedingly cheap. 8-bay eSATA RAID cases run a couple hundred bucks, new. 4TB SATA drives run less than that. With 6 raid ports on a mainboard and a couple of dual- or quad-eSATA port PCI-x cards, you can approach petabytes quickly—and just for four digits. The same goes for processing power—a dual-processor Xeon setup (in which each processor can have core counts in the double digits) again just runs $couple thou.
And data is now cheap and easy. Whatever you want—you can have it as data *already*. Movies? Music? Books? Big social data sets? They're coming out our ears. The investment of time and equipment required, all in all, to put yourself in a position to rip and store a library of "every movie you've ever rented," and then actually do so, is much larger than the cost of simply licensing them via streaming. The same goes for music, ebooks, and so on.
There's just no need. Even my desktop is now starting to feel obsolete—for the work computing I do, there's a good chance I'll just go to Amazon cloud services in the next year or two. At that point, an iPad, a wireless keyboard, and a couple apps will probably be all the computing power I need under my own roof. If I have a desktop, it'll just be to connect multiple monitors for screen real estate.