I presume Vadim refers to the peer-to-peer component of WebRTC, not the camera access component.
To answer Vadim, it's probably possible to read the image data (since I saw demos where the image can be painted into a canvas), but it would almost certainly not be as efficient. For one, you'd at the very least have to implement a codec in Javascript to avoid uploading uncompressed image data, since you can't get just the compressed stream for further transmission via your own channels.
In short, you'd be better off using WebRTC on the client side, and figure out the best way to write a server side component. One way is using Chrome's open source WebRTC libraries, but you could also just implement ICE and RTP differently, and parse SDP on your own.