My first thought was "Black has elected to play against the Blackmar-Diemer a move down?" I'm firing off the top of my head here, with little time for deep analysis.
It's an interesting position, but I can't help but think it's already lost for Black, somehow. White can throw away a move in the opening, Black shouldn't be able to. With Black throwing away a move like that my instincts are to go for rapid development and open lines, when the lack of a move will be the most painful.
My first thought would be Be3, electing to actually play the gambit if Black finally accepts the pawn. But that may not be to your taste (if you don't know the typical positions and tactics, it can be dangerous). I'm not fond of Bd3 here, for some reason. If I just wanted to get out of this into normal, basically equal waters I might try capturing the d-pawn and playing out the exchanges.
Bg5 on the surface looks interesting. You'll need to beware of the possible check-and-fork (Qa5) but your threat to win the pawn outright should net you a very weak Black pawn on e4 that White can play against at leisure, probably eventually winning it (Nge2-g3 comes quickly to mind as one way to play for it -- your opponent has thrown away a move, so you shouldn't feel too badly about taking two to place your Knight on a useful square). One advantage of Bg5 on this is it clears the way quickly to castle long, when you'll either be able to also bring Queen and Rook to bear on the weak e-pawn (if your opponent chooses that path) or have your king in safety as you rip open the center and go hunting for the enemy king.
(Some wild lines are dancing through my head about this; you might end up in positions similar to the wilder side of the semi-slav but with having played Nc3 instead of c4.)