Excuse me as *coughbullshitcough* I clear my *ahembullshitahem* throat.
50x is not possible. He's claiming 50x faster than Bittorrent at DVD quality. Not possible through a protocol.
Assume that you've got connection and bandwidth parity - remember, we're talking about the protocol here - for whatever he's talking about (Blin) and for BT. He is then saying that the bittorrent protocol adds in overhead that slows it down by 50x, comparatively. That's total BS.
You can't factor data compression in (as that doesn't talk about the protocol) because it's already video and audio compressed (if AC3 or DTS and not LPCM). Data compression doesn't significantly reduce the size of compressed audio or video in most cases. So data compression is out even if someone wants to try to factor it in - it's irrelevant. Remember... DVD quality... (Lossey compression degrades quality.)
What's left? Overhead? Nothing is that bad. You would have to TRY to make the protocol as horrible as possible. Only government is capable of doing something that badly.

The ONLY thing that makes real sense is that in China fewer people use BT while many use the other program (Blin). More clients means more places to download from. THAT can result in faster speeds like 50x, but not a protocol.
Here's some simple math:
1 DVD = about 4GB (4,294,967,296 bytes)
Running time = about 2 hours
Download speed is then 596,523 bytes per second
Which is entirely believable (I get MSDN downloads at much faster speed than that), but it also says that BT speeds are then 11,930 bytes per second (as it's 50x slower), which is true in some cases, but it's not hard to get a few hundred thousand bytes per second with BT.
That the protocol is 50x faster is simply false.
That the Blin P2P network is much larger than the BT network in China is entirely believable. Those are VERY different things.