If Qd7 is not played, then Nb5, which cuts off our queen's protection and essentially forces a trade, is strong for white.
Even if Qd7 does not bring an advantage, it may be the only (or best) way to avoid a disadvantage.
From a strategic viewpoint, an equal position may be best in order to try to get Martin to overextend. If he doesn't, however, then a draw becomes a likely outcome.
In the end, it is better to have a position where we are forcing his moves but have options ourselves. This way we can focus all of our attention on one variation, whereas Martin has to look at several.
Of course, an equal or better (if only slightly) position is better than a worse one.
e5 may not work.
Nxd4 e5
Nb5! Qxb3 (forced)
axb3 Bxc1 (and he can take with his King or his rook; very dicey because he can take our LSDB with his rook, next move, and who knows the havoc that may follow… but assume he doesn't)
Rxc1
We cannot castle because Martin would play Nc7, forking two rooks and a DB all at once.
Now, instead of castling long, we could capture the knight with our DB (a gambit of sorts) and he captures back with his pawn, thereby doubling his pawns on the b file.
Haven't worked it out since I did all of this in my head with no board in front of me.
But e5 has serious ramifications that have to be considered.
We would most likely castle long next, and he has to castle long, too.
We bring our knight out to d7 (have to or we are doomed), and he plays a rook to d1.
So, it is the doubling of his pawns versus our poor trade of a LSDB for his knight.
It might be advantageous for us. Don't know.