Refer to this: https://www.gnu.org/philosophy/free-software-for-freedom.htm...
At a high layer of abstraction, neural connections work probabilistically, firing in the path most likely to be a useful pattern; hence, most people will tend to assume "free" means "as in beer". However, a new term like "liberated software" will not trip this pre-existing wiring, forcing a new pathway that doesn't have to compete with the conditioning created by a lifetime of marketing messages ("Free Checking! Free Estimate! Free For The First Year!"... etc)
Good thing 4-clause BSD never caught on in the Linux world.
By the way, have you heard about Freenet? https://freenetproject.org/
You said it downloads the entire site when i open a site. Now, this is completely fine for normal text-based blogs, but what if it's an image heavy site? Like theChive? Wouldn't i end up storing several gigabits (possibly terrabytes) of data for that site?
Secondly, if somehow we extend the concept to dynamic content (don't ask me how) wouldn't it effectively mean each of us running a server and hosting it on that?
Maybe your idea can be implemented on the server level instead of the client level... like proxy servers doing what you want the clients to do and thus the content would be all over the internet instead of a few specific servers. (not saying that's a good idea btw)
The biggest drawback is that you can only serve static content, although I think many websites today would be fine with this restriction.
You can read more on my website: http://jack.minardi.org/software/syncnet-a-decentralized-web...
Follow along with the development on github here: https://github.com/jminardi/syncnet