It allows your opponent to develop pieces faster and have control of the centre.

There queen isn't developed too early either. Because you traded off the knight that would normally attack it, it's harder to get out of the center

Just look at the resulting position.  White has a central pawn and one piece developed. Black has nothing developed, though it is black's move.  With no convenient way to harass the white queen, black's development is in some ways actually more difficult due to the queen on d4 (as opposed to other early queen moves which aid development) -- for example, ...Nf6 can be met with e5, and ...g6 to develop the dark-squared bishop obviously just loses the rook.

You can take here - it's not too bad. You are swapping a knight on the third rank for one in the center, so you're getting the better deal. Obvioulsy the down side is that you develop his queen but you may be able to gain a move by attacking it with a supported Bc5 or e5.

Black cannot easily develop and is falling rapidly behind.  Nf6 is met with e5 leaving a choice of retreating or Nh5. Queen moves offer little and all other developing moves require preparation thus Black is down a couple tempi.

In this opening, sometimes Black doesn't get developed fast enough and the king is caught in the center.  Here is a game I played with this opening.

Wow. Wonderful instructive game Bill. I've been looking at this and although i haven't checked with a computer I think you played very accurately. 8.e5!! is a move that would never in a million years occur to me -- I would avoid it thinking that it encourages an exchange of queens which wouldn't help white, and it seems to set up a discovered attack on the white queen along the long diagonal.   But of course that discovered attack I'm so afraid of works both ways (and is indeed the source of the winning tactic). And e5 follows the attacking principle:  when his king is in the center and your's is castled, open lines in the center.  So after the forced (right?) 8...de 9.Qxe5 Black plays the very very natural 9....Bg7? and from there I don't see how he saves his game! 10.Rd1! (how easy it would have been to miss the winning attack with some bland 'developing' move. For example 10.Nc3 0-0 and the game is equal I think.) And 10...Qb6 is forced -- the black queen's only square!  11.Be3 develops and the harassed Black queen has just two squares -- he plays the more natural one Qe6... but Qb4 would also fail to the same combination as played in the game. 12.Bb5+ and black has no good answer....

A forced win from move 9 and black plays only natural moves with no obvious blunders?? This is one helluva trap, if I were a 1.e4 player I'd memorize this thing today. 

Well, actually, first I'd check my analysis with a computer, because I probably missed something!

How far back does Black have to rewind to get a playable game? Maybe 9...Qd5 rescues black? 

I'd hate to sound cynical but basically all the way to 4. ... Nxd4. There's no need to exchange the knights, it doesn't gain anything and wastes the tempo black has already invested in the knight. If white captures the knight, black can simply recapture with the b-pawn and have his center intact. Just my few pennies.

edit: after taking a closer look and putting some thought into it, I think 5. ... d6 avoids what transpired in the actual game as would've 7. ... e5. Sure, black has that backward pawn afterwards but at least it's not the trainwreck it was after 7. ... g6 8. e5. And finally, moving the queen to b6 is what tipped everything over. Blocking the rook with the queen's bishop seems plausible.

I usually respond with e5 instead of Nxd4.
Most trouble from there comes with the weak d6 square, but never something beyond fixing, at least that's my experience with it.

As far as the the little opening theory I know goes, the main problem of the open sicilian seems to be in fact the d5 square. :)

You're probably right, I never know the exact lines. But what I meant was after e5 Nb5 gets queen and knight attacking d6. This I usually solve by d6, but I'm really not sure whether this is the correct way either. But I do know I don't like the knight on d6 with a +

In my game, I think 9...Bg7 is playable, but 9...Qd5 may be better (9...Qd5 10.Qxd5 Nxd5 11.Rd1 Nf6 12.Nc3 and 13.Be3).   After 9...Bg7 10.Rd1, Black can play 10...Bd7, then castle.  I think 10...Qb6 is the losing move as Black cannot prevent 11.Bb5+.

I don't think Bg7 is all that playable, although you're right I missed the Bd7 block... but check this out:

Ah yes, I see your point now.

Well, the way I see it it's either lose a queen and castle or run amok with the king. I don't see white threatening an immediate mate against correct play, but yeah, the situation is way out of my comfort zone.

Other option is to take with knight, let white have the bishop, bail out and castle queenside.

OR just avoid an unnecessary knight trade in the center to begin with. It just wastes time and if your knight is captured, you still have your pawn structure intact and white has just wasted a tempo helping you to achieve that and the semi-open b-file :)

I think Black should be perfectly ok with 7..e6 followed by Be7 and 0-0.

 7...e6 certainly looks playable -- but that means black abandons the kingside fianchetto idea... which in the sicilian feel like a bit of a concession right there, doesn't it? 

You know, I think billwall's game goes a long way toward explaining why 4...Nxd4 isn't recommended! (yes, poidokoff, I know you've been saying that all along. ;))