An interesting game in the 2....f6 variation was:

[Event "Kastav Hotel Opatija-A 04"]
[Site "Kastav"]
[Date "2002.04.10"]
[Round "9"]
[White "Mukic,Josko"]
[Black "Ribli,Zoltan"]
[Result "0-1"]
[Eco "D00"]
1.d4 d5 2.Bg5 f6 3.Bh4 Nh6 4.Nf3 Nf5 5.Bg3 h5 6.Bf4 g5 7.Bc1 g4 8.Ng1 c5 
9.e4 dxe4 10.d5 e6 11.Nc3 exd5 12.Qxd5 Nd4 13.Qxe4+ Qe7 14.Bd3 Nbc6 15.Be3 Be6 16.Nd5 Bxd5 
17.Qxd5 Ne5 18.Bxd4 Nf3+ 0-1

One aspect of 1 d4 d5 2 Bg5 f6 3 Bf4 is that black could now get a whacko Blackmar-Diemer [without being a tempo behind - only with differently coloured pieces] by 3 - e5 4 dxe5 Nc6 and white will now find himself in one of the 'trickier' lines of that gambit  ...  in fact, native wit and common sense might well not compensate white for lack of knowledge of the 'tricks' in this complicated variation. The rate at which the "untheoretical" Trompovskies are becoming theoretical seems to be accelerating. doesn't it !
[ PS ... I think Eric's site is one of the very best on ChessPub - loads of originality and honest work]

This is indeed a bigger problem, than the well-spread 2...c6 or 3...h6 4.Bh4 c6 on which will focus the next update as well as on the Tromp 2...d5. 
Much bigger than I thought at first sight when  I decided to play the Pseudo-Tromp, as Hodgson himself has started to play 3.Bf4 instead of 3.Bh4...
Actually, I am not satisfied with what I have seen in the database. So I guess, I shall have to study it properly over the board and look for some new idea.
I will let you know