You won't come anywhere near a 10x compression ratio. I reckon you'd rather see really some low 1.x factor in real world scenarios. Also keep in mind that a VPN's encryption adds a bit of overhead.
What your article points out is a way of getting rid of packet loss in wireless networks / speeding up retransmission and thereby getting closer to the maximum packet rate possible. It also points out at the beginning, that packet loss usually is not a big issue in wired network connections. Also it is stated, that "On a loss-free network, coded TCP won’t do a thing; this innovation isn’t suddenly going to boost the throughput of 802.11n networks from 300Mbps to 3,000Mbps.".
It is just a marginal amount of packet transfers that can be sped up 10 times by this technology, it's not the whole connection speed that's boosted. On a 100 mbps connection, you'd still get a maximum of 100 million bits per second in each direction.
If there was an easy way to compress traffic by ten times we'd all be doing it for years already. But how would you compress for example a packet, containing a part of a jpg or mp3 (not that you'd know about the content, anyway) by 10 times!?
If you knew, I bet you could make a fortune
Depending on the scenario you'd be better off with some proxy server, output compression on web servers, etc.