Why does white fianchetto against the Dutch?

By the way, that's a nice idea Thrillerfan, but I don't think it's the whole story. You actually (almost) never see the queenside fianchetto covered for black, and for good reasons. If white plays a c4-d5 centre, then actually the bishop accomplishes nothing on b7, where as it is freed automatically as soon as white plays e4! If black manages to get the centre with e5 then the common black advance with e4 boxes the bishop back in again! Only the other typical attack with f4! from black releases the bishop, provided that it is on c8 or d7. That means the fianchetto will only work in some limited cases where the centre is fairly open.

d5 is not that easy for White to achieve against the Fianchetto setup and be able to hold on to it.  Keep in mind, you don't double-fianchetto.  If White allows b6 and Bb7, which the Bishop also covers d5, you don't then fianchetto the Kingside Bishop also, you play e6 (which also covers d5).  The White Light-Squared Bishop, don't forget, is not fianchettoed as Black beat White to the punch in this case.

And actually, many books DO cover the b6/Bb7 idea against the lines where White plays e3 instead of g3.  Case in point, Win with the Stonewall Dutch, covers it against many of the sidelines, such as say, 2.Bf4.

So with the Bishop on b7, Pawn on e6, and Knight on f6, all covering d5, with White unable to get his Bishop to eye d5 all that easily (if would have to be on the a2-g8 diagonal, if it goes to f3, Black can trade it before White gets in d5), you are relying on c4, a Knight on c3, and a Queen to cover.  Of course, that Knight on c3 can often be eliminated by Bb4.  Keep in mind, this ain't a stonewall or classical.  No reason to hold the DSB.  If anything, this line plays a lot like the Dutch Variation of the Nimzo-Indian Defense, so Bb4 is actually very appropriate!

Your counter-argument against playing b6 and Bb7 when White allows it is extremely weak!

Because black can mate in early moves in Dutch so that is why the white wants a guy near his king.

You're talking about some very specific cases here, for example when black has already been committed to e6. Then after e3, you might go b6, Bb7 and the other bishop sometimes goes to e7, sometimes goes to b4. I should add ''might,'' because you really don't have to. You can also follow up with e6 instead. Against the Leningrad it really doesn't matter either if white plays e3: Black should be fine anyway.

Unless you have a pretty specific book for the old Dutch or the Stonewall, the fianchetto is not going to get covered. In fact I only found a reference to it in my old opening book from Euwe. 

Williams with his "Killer Dutch" DVD   (Classical Dutch)  is very keen on this fianchetto for black with Nimzo like play against the "minor" lines white can play.

Yea I'll just keep posting master games until you decide that you're okay with how black played the position.  That way you can continue to claim it's a bad system with fantastic analysis like,
"It's just common sense."

@ThrillerFan
"It's to PREVENT BLACK FROM FIANCHETTOING THE LIGHT-SQUARED BISHOP!"
That's one idea but my database has 14 games with black achieving a queenside fianchetto and white won most of them. It's not the losses but the very mediocre position for black that makes me think your theory is incorrect.

In the position above with black knight on c6 and white Knight still at home, as black I would have liked to develop my g8 knight to f6 first. IF white develops Nc3 then the nimzo idea as well as Ne4 and nxe4 and then black playing Bxe4 "pinning" white's knight to the white's LSB. I would, as black prefer developing the b8 knight to d7....

First off, the lines I advocate would NEVER match what you have listed there.  The Knight has no business on c6 that early.

Also, Nf6 should be played, not e6, on move 2.

I'm talking secenarios like the following:

1.d4 f5 2.c4 Nf6 3.Nc3 e6 (to prevent d5) 4.e3 b6

1.d4 f5 2.Bc4 Nf6 3.e3 e6 4.Nf3 b6 (London players have done this - It's known as the Dangerfield Attack, mainly because it gets no respect)

1.d4 f5 2.c4 Nf6 3.Nc3 e6 4.Nf3 Bb4! (Play is very similar to the Nimzo-Indian)

1.d4 f5 2.Nf3 Nf6 3.e3 e6 4.c3 (Colle players have done this)

1.d4 f5 2.c4 Nf6 3.Nc3 e6 4.Bg5 (A very amateurish move) Be7 and if White doesn't play g3 now, b6 is coming.

There may be others, but this is just to name a few.  Notice how all of the lines I have listed have one of two features.  Either A) e3 has already been played, and so e3 combined with g3 is very weak for White, especially the Light squares, or B) White has yet to develop his King's Knight, like in the last line mentioned, and so White will never get what you have shown.  If White answers b6 with g3 in a case where e3 has not been played yet (like the last line), then after Bb7, White will have to play Nf3 and Black will pop it and wreck the pawns, followed by c6, similar to the English Defense.  If White doesn't play g3 then, when is he ever?  Your Knight is NOT going to c6.

With me it would depend on White's behavior.  There are a few rare instances where the Knight might actually be developed to a6.  Again, rare, but possible.

There are actually many instances in the Leningrad Dutch where the knight can be advantageously developed to a6, either before or after a5.  The point is to control the dark squares on the queenside.  If all fails, c6 and Nc7 provides a good spot for the knight while covering the vulnerable d5 square.

 I thought Thriller was trying to say that once black got in the queenside fianchetto then black is positionally won but I see I totally misunderstood that.

Agree with the above post about preventing Bb7. But I'd also add the the Bg2 is a useful defender of the king if black tries to launch an early k-side attack. Without the k-side fianchetto, moves like Bd6, Qh5, Ng5 etc can be quite annoying in some positions.

I like using the Dutch if I can tell an opponent hasn't seen it before, because if they don't fianchetto the kingside bishop I can usually smash them with a quick attack.  I do not like fianchettoing the kingside bishop myself as black, because I feel it leaves the kingside too weak.  If I play against the Dutch as white, I always get the kingside fianchetto in, because it's so hard to break through on a side with a fianchettoed bishop.  Then I just wait for weaknesses to pop up and I exploit them.  

Didn't Alekhine used to fianchetto his lightsquared bishop in the Dutch?  The problem I see with it is a well-timed d5! strike from white can leave the e6 pawn rather weak (dxe6,dxe6)

Another benefit of a kingside fianchetto from white is pressure against the b7 pawn, which can become quite weak in the Leningrad:

It was a remarkable and quite a high quality accurate game!  Nakamura's defensive technique was simply splended!  He could have gone wrong at so many points but didn't.