4.e4... what KINDERMANN is telling about this!?

Greetings

Well, there's exactly one game in my database after 4.c3:
Kolvig,B - Moller,P [A80]
DEN-ch Nykobing (2), 1961
1.d4 f5 2.Bg5 g6 3.Nd2 Bg7 4.c3 d6 5.e4 h6 6.Be3 Nf6 7.exf5 Bxf5 8.Qb3 Qc8 9.Bc4 e6 10.h3 d5 11.Be2 g5 12.g4 Bg6 13.h4 gxh4 14.Rxh4 Nbd7 15.Ngf3 c6 16.0-0-0 Qc7 17.Rdh1 0-0-0 18.Qa3 Kb8 19.Ne1 Bf8 20.b4 Ne4 21.Nxe4 Bxe4 22.f3 Bg6 23.Bd3 Bxd3 24.Nxd3 Bd6 25.Rxh6 Rxh6 26.Rxh6 Nb6 27.Qb2 Nc4 28.Qe2 a5 29.Kc2 Qb6 30.Kb3 axb4 31.cxb4 Qa6 32.a4 Nxe3 33.Qxe3 Qc4+ 34.Ka3 Qc3+ 35.Ka2 Bxb4 36.Rxe6 Qc4+ 37.Kb1 Bc3 38.Nb2 Qb3 39.Qc1 Rh8 40.Re2 Ka7 41.Qd1 Qb4 42.Rc2 Bxd4 43.Ka2 c5 44.g5 Re8 45.Rg2 Re1 46.Qc2 c4 47.g6 c3 48.Nd1 Qc4+ 49.Kb1 Qf1 0-1

Then via different mover orders there are a number of games after 5.e4, mostly played by people rated around 1700 (oops, shockingly enough even two of my own blitz games - where I won handily as white - fascinating)with the exception of the Ragonese-Saccone game I already quoted (Saccone had an Elo 2175 at the time) and the following desaster (where no rating for black is given):
Milisits,R (2215) - Iinuma,P [A80]
US Open Los Angeles USA (12), 15.08.2003

1.d4 f5 2.Bg5 g6 3.Nd2 Bg7 4.e4 d6 5.c3 fxe4 6.Nxe4 Nf6 7.Bd3 0-0 8.Qe2 Nc6 9.0-0-0 Qe8 10.h4 e5 11.Bc4+ Kh8 12.Bxf6 Bxf6 13.d5 Ne7 14.g4 Bg7 15.Nh3 h6 16.f3 Qa4 17.Bb3 Qa5 18.Qd3 Bd7 19.h5 gxh5 20.Neg5 hxg5 21.Nxg5 Nf5 22.Rxh5+ Kg8 23.Ne6 Qb6 24.gxf5 Bf6 25.Qe4 Kf7 26.Rh7+ Ke8 1-0

And indeed, quite possibly my 5...h6 is misguided (if I want to play h7-h6 it's probably better one move earlier, I suppose).

But basically I don't think there really are any games by really strong players and the best I've found are the recommendations by Kindermann (who does suggest 4...d6, but ignores 5.e4) and by Pedersen (who does suggest 4...h6, but also ignores white's option of e2-e4). An alternative suggestion by Pedersen is the 3.Nh6 that TalJechin also mentions, but so far I really like the standard mainline after 3...Bg7 4.e4, so I'd prefer to repair that.

Don't know much about the theoretical status, but 3...Nh6 intending Nf7 á la Staunton, is an interesting way to play it for black imo.

I have recently lost the following rather horrible game:
1.d4 f5 2.Bg5 g6 3.Nd2 Bg7 4.c3 Well, I didn't really know anything about this variation, so I was making up everything over the board from here onwards... 4...d6 My move is even what Kindermann recommends, hower he does not even look at white's next move.
http://www.france-echecs.com/diagramme/imgboard.phpfen=rnbqkxnr%2Fpppxpxbp%2Fxxx...
Somehow Kindermann seems to assume that white has to play with c3+e3. But why? He can just play 5.e4 . Maybe Pedersen's recommendation of 4...h6 is better? One move 5 white's bishop would presumably have to go back to f4.
http://www.france-echecs.com/diagramme/imgboard.phpfen=rnbqkxnr%2Fpppxpxbp%2Fxxx...
Doesn't 5.e4 seems completely logical? Still it's not covered by Kindermann.
5...h6 6.Be3 Nf6 7.Bd3 e5 Maybe rather something careful like 7...e6?!
8.exf5 gxf5 With illusions about getting dynamic play using my central pawns. Maybe 8...Bxf5 9.Qc2 Qd7 would have been a more sensible idea? Now I might even be able to consider Bxd3, as I will have Qd7-f7. 9.dxe5 dxe5 10.Qc2 e4 
http://www.france-echecs.com/diagramme/imgboard.phpfen=rnbqkxxr%2Fpppxxxbx%2Fxxx...
My suspicion is that here I'm already lost.
11.Be2! 
Okay, so this move by my opponent looks like a massive improvement over 11.Bc4 Nc6 12.0-0-0 Qe7 13.Nh3 Be6 14.Qb3 Nd8 15.Nf4 Bxc4 16.Qxc4 Qf7 17.Qb5+ Nd7 18.Nc4 a6 19.Qd5 Qxd5 20.Rxd5 ½-½ Ragonese,L-Saccone,F/Vitinia 1996/EXT 2000 (20), which seems altogether fine for black (to me). 11.Be2 seems much better at exposing the positional flaws of black's position.
11...Be6 12.Nh3 Nbd7 13.Nf4 Bf7 14.g4 fxg4 15.Nxe4 Ne5 16.Rd1 Qc8 17.Qa4+ Kf8 18.Ng3 h5 19.Qa5 Nc6 20.Qg5 Ne8 21.Bc5+ Nd6 22.Nf5 Be5 23.Rxd6 cxd6 24.Bxd6+ Bxd6 25.Qg7+ Ke8 26.Nxd6+ Ke7 27.Nxc8+ Raxc8 28.Ng6+ 1-0
I'd be quite interested in thoughts about the opening phase. What is black's most solid/respectable way of playing here?