I think the key point is the effienctly compute in the defintion. It seems that the black box property means C() and an oracle version of A() must be completely indistinguishable, including things like runtime and memory usage.
The weaker obfuscator has no guarantee about that.
I'd argue that bittorrent is still better for unpopular data then direct downloads.
I think BT makes downloading and sharing accessible enough that more people would be encouraged to download more than they normally would and it is for sure easier for someone to see a request for a reseed for an unpopular torrent and load it up onto their tracker for a few hours than it is to try and rehost it.
Surely if one other person is downloading the same torrent at the same time it's better (potentially) than a direct download, as part of the torrent can be shared in the swarm..
As long as archive.org is seeding everything all the bits are there..
This article seems confusing, they are using the browser specific implementations of CSS3 and passing it off as the standard naming (eg. box-shadow vs -webkit-box-shadow) without saying anything about it.
That's just javascript minimization. Standard way of reducing page load times. It's been around way longer than Google Closure, just look at YUI Compressor.
It may stop anyone from looking at a page's source code, but I'd think most developers would put greatly reducing load times ahead of having your javascript being freely read.
The weaker obfuscator has no guarantee about that.