Btw, I dont like ...c7-c5 in these kind of positions. Maybe this is a matter of preference but in general I have more faith in ...c7-c6.

(1.d4 Nf6 2.Nf3 d6 3.c4 Nc6 4.Nc3 e5 5.d5 Ne7 6.e4 Ng6 7.g3 Be7)
I suggest:
8.Be2 c6 9.h4 cxd5 10.cxd5 Nf8 or
8.h4 Nf8 9.Be2 Nfd7 10.Be3 0-0 11.Nd2 (Rohde-Hagesaether) and here perhaps 11...c6 again.
I am avoiding 8.Be2 c6 9.h4 Nf8 10.c5 or 8.h4 Nf8 9.Be2 c6 10.c5 (maybe not deadly, but still annoying).

I still haven't had a thorough look at the positions arising after 6...Nd7 but I must say I like it less and less. Not so much because of 7.h4 (which btw may be a decent try as well) but mainly because of 7.Be3 and even 7.g4!? Im not sure about the ladder but 7.Be3 certainly feels right. The bishop no longer has to worry about ...Ng4 ideas and also supports a quick c4-c5 thrust which must be a critical test to Blacks setup with knights stuck on d7 and e7. 

One should also note, that having both kings in the center almost unvariably favors White, because his space advantage allows him to more freely coordinate actions on both flanks whereas Black does better to just castle and quickly focus his attention on one of the flanks (usually the kingside but not necessarily).
In this opening, I can hardly think of any likely scenarios in which Black can seriously consider ...0-0-0 nor just leaving his king in the center, whereas White may choose either of the three options. Therefore, Black does better to just quickly castle and start playing on one of the flanks.

Yeah, you are probably right h5-h6 is very double-edged. But I just thought that h4 is more forcing (and thus more precise?) as all other moves than 7...Be7 seem to be bad. My idea was to tempt Black to play Bg4/h6/h5 neither of which I think are very strong. One should also note that conceptionally it may be harder for the Black player to correctly forsee and evaluate 7...Be7 followed by 8...Nf8. If White plays 7.g3 then 7...Be7 is by far the most logical response and after 8.h4 then 8...Nf8 is somewhat easier to find and evaluate. So my point is just that 7.h4 opens up to more pitfalls for the Black player, although if the Black player knows/sees the plan ...Be7 followed by ...Nf8 then this is just a transpositional matter.


On another note, I couldnt help wonder how White is supposed to punish:

1.d4 Nf6 2.c4 Nc6 3.Nc3 e5 4.d5 Ne7 5.e4 d6 6.Nf3 Nd7!?

If Black doesnt trust the Panther variation isnt this just a viable way for Black to reach a KID?

Correct me if Im wrong but isnt 7.h4 just an improvement over 7.g3 ?

OK, I see what you mean. 
1.d4 Nf6 2.Nf3 d6 3.c4 Nc6 4.Nc3 e5 5.d5 Ne7 6.e4 Ng6 7.g3 Be7 8.h4 Nf8 9.Bh3 Bxh3 10.Rxh3 N8d7  11.h5 c6)
Yes, 12.Nh4 g6 13.hxg6 hxg6 followed by cxd5 or 12...cxd5 first.
Or 12.h6 g6 followed by ...cxd5.
Black gets the c-file and White's pieces are starting to look a bit uncoordinated to me.

[quote author=2122282E4F0 link=1233890161/20#20 date=1293782561]A couple of years on it looks like there hasn't been much action in this theoretical backwater -- what a surprise! Taking another look at some of the above lines, my attention fell on 1 d4 Nf6 2 Nf3 d6 3 c4 Nc6 4 Nc3 e5 5 d5 Ne7 6 e4 Ng6 7 g3 Be7 8 h4 Nf8 9 Bh3 Bh3 10 Rh3 Qc8 11 Rh1 N8d7 12 h5, when TN gives [u]12 ...h6[/u] 13 Nh4 Nc5 14 Qe2! as good for White. Instead of 12 ...h6, I've a dim memory of having looked at and rejected the obvious [u]12 ...Nc5[/u], but I can't now see why. Is h5-h6 really a threat? How about 12 ...Nc5 13 Qe2 a5, meeting 14 h6 with 14 ...g6 and then ...0-0? Engines seem to think this is reasonable, but then it's a very unengineish position, so to say ...[/quote]
Interesting, but it still seems like Black doesn't have much play and White has a lot of time to rearrange his pieces. White still looks better.
Going back a bit, I don't really understand the purpose of 10...Qc8. I notice that TN wrote earlier:
[quote author=407C71446671677160677B7226252424140 link=1233890161/13#13 date=1234094493]
c) 8...Nf8 is possible, but now White is up two full tempi compared to the 5.e4 Be7 6.g3 Nbd7 7.h4 Nf8 Czech Benoni - no ...c5 and Nf3 included for White. This has to give him an advantage, e.g. 9.Bh3 Bh3 10.Rh3 N8d7 11.h5 and even though the computer claims equality, I think White is better due to his extra space and slightly more active pieces.
[/quote]
I wonder. The extra ...c5 in the Czech Benoni means no ...c6, and the extra Nf3 means the f-pawn is blocked. 9.Bh3 Bxh3 10.Rxh3 N8d7 11.h5 c6 and I think Black is doing well enough here.