How many drops in the cloud?

From this post on data management:

Data also has to be accessible and stored securely. 56% of the researchers surveyed stored their data on a hard drive. There is a problem with storing data in one place, as many learned the hard way from Hurricane Katrina, when an incredible amount of data was lost. ... Noble suggests storing data online, as hard drives cannot be accessed once you are out of the lab, and as another researcher pointed out “I think there should be online storage of research data … so when ever you have time you can analyze your data.”

There’s a danger though of thinking that data stored “online” is somehow different than data stored on your desktop computer’s hard drive. It isn’t. That data is still dependent in some way, shape, or form, on being stored someplace on a physical object. Data on the web are not like Cartesian souls, immaterial information free to exist independently of the physical world.

If one of the platters spinning at Blogger or Google malfunctions or is destroyed that contains the archive of this blog, is there another copy to restore it? I don’t know. I haven’t heard any horror stories yet, but that I am ignorant of the answer is worrisome.