Caro Kann with 3.d3!? What do you think?

I discovered this line thanks to GM Gascon del  Nogal on his Youtube channel. If you don´t want to study a lot or are bored of the main CK lines, this looks like an excellent choice.

I've won a lot of speed games with 1.e4 followed by 2.d3 against anything.

French and Caro players sometimes trade and take your queen... I guess thinking that at worst it's just equal for them.

Then I often get ridiculous scores like +1 from the engine by move 15. Stick the king on c2, advance pawns on both sides. Black needs to realize his pawn break is f5... that's not the only way for black to play, but typically even when the position remains equal they burn tons of time trying to figure out what's going on.

I've played against it too, as black, and it can be pretty unpleasant in speed games. Sometimes it's just a lot easier for white to find moves.

Not as simple as it looks.  I think you've got some trouble on your hands after black plays 7. ..Nbd7!, threatening Ne5, and if 8. h3 Bxf3 9. gxf3 Ne5! 10. Be2 O-O-O+! with a strong attack.  On all lines, the pin on f3 seems to cost white tempo.

You are right, that looks like the most promising answer for black. Even so, it is playable for white. 8.h3 looks premature. 8.Nbd2, developing and defending f3 should be better.

Playable yes, but I'd still prefer black's position.  After 8. Nbd2, which I agree looks best, 8. ..Ne5! hits the still-pinned knight on f3 a second time, along with the d3 bishop.  Black's knights start to look pretty menacing, as does the threat of a queenside castle.

If you don´t like dealing with the pin, you can follow the amazing Anish Giri´s steps, moving four times the knight in the first nine moves...

I think the exchange 3. ..dxe4 followed by the trade of queens is clearly best for black.  Giri's 6. Nfd2 seems to be necessary, there Nepo's reply (6. ..Ng4) is questionable.  Black should continue development and prepare to castle queenside, especially with black's dark square bishop hemmed in like that.

This seems like a decent off-beat try for white. I mean black can still equalize, but it could be dangerous as a surprise weapon.

I much prefer 2. d3 followed by 3. Nd2 to prevent the exchange of Queens.

Then I can play it like a King's Indian Defense with the colors reversed. According to the data-bases, White already has a 66.5%-to-33.5% advantage by move 6. The book moves (after 6. ... d4) are 7. Be2, 7. g3 or 7. Qg3... but personally I would think that 7. a3 (to prevent Bl;ack from trading that bad Bishp at f8) or 7. g4 would be more to the point.

After 7. Qg3 there is 7... Nd7 8. Nc4 f6 to defend the pawn, but it doesnt look good for black

I think after 1. e4 c6 2. d3 d5 3. Nd2 e5 4. Ngf3, 4... Bg4 is inacurrate, 4... Bd6 is better.

If I had the Black pieces, I would just continue developing, and ignore the offer to exchange queens. It would lead to an interesting game, but there's really nothing in it that would concern Black.

For example:

Of, course, is not a line which gives white a big advantage, but it is interesting anyway.

I play the CK frequently and have faced 2. d3 often in speed games.   It's not a problem as it is quite passive and gives black multiple options.  I also have gone the trading Q's route but now I usually avoid a Q trade unless I can do more than prevent W from castling.   It might work well on players not familiar with the CK but are playing it anyway.

Yeah, it's definitely interesting.

I remember seeing a simul game of Kasparov's (from a book collection), in which his opponent offered an early queen exchange, in a similar manner.

But Kasparov ignored the offer and continued developing, instead. I thought it was a surprising choice, and perhaps an instructive one, too.

Kasparov comes from the Soviet school, which teaches that completion of development should precede any tactical efforts in the opening.

In this particular situation, white can't play exd5 without disadvantage (the cxd5 recapture gives black a central majority, frees the knight's main mobilization square, and leaves him a convenient open c-file), so to a certain extent, it stands to reason that is to black's advantage to leave the positional tension in place until a more complete plan of attack has been prepared.

That's a reasonable way of thinking. One thing you can do to encourage this sort of sequence as black is look for ways to pressure e4, which may cajole white into capturing.

However as you said, white doesn't want to lose the central majority, so he's more likely to advance the pawn or fight hard defending e4. Ultimately if you can't find ways to pressure e4 then the whole plan may be a pipe dream.

Attacking the e4 square (and as Nimzowitsch explained, it is the square that is attacked, whether there's an enemy piece on it or not) is the basic premise of the Caro-Kann, so pressuring that square, and/or setting up a blockade on e5, would be very much in keeping with the reasons behind playing 1. ..c6 in the first place.

Not every attack carries (most of them don't), but pressure on an opposing weakness in itself can generate activity elsewhere on the board, and prevent opposing pieces from becoming active.