While we are talking about the Caro, when I still played 1.e4 I used to play the Panov attack vrs the caro-kahn
1.e4 c6
2.d4 d5
3. exd5 cxd5
4. c4
A very stong club player claims that this is refuted by 4... e5. Personally I don't see it though. My database contains 2005 games of Topalov playing the Panov attack, so it's hard to consider that anything but a joke, but how would you all continue against 4... e5? The chessbase big database contined 2 games with that position, both from 1960 or earlier.
Has anyone seen the following line?
e4 c6
Nc3 d5
Qf3!?
I had a brief look at it when I was considering playing e4 as well as d4. It has an advantage of being light on theory, although admittedly black can equalise with accurate play (it does hold a trap or two, however). Of course, with it being a rare opening, white could also have an advantage of being more familiar with the positions and formations arising from the opening.