Is this really a Queen's Gambit?

Not normally, no. That's an anomalous square for the Knight in the Najdorf. Nd7 is the usual spot, and the d7-Knight is interested in a different constellation of squares than a c6-Knight.

Then, why is this called a Najdorf?

Because it would not have arisen from the 2. ... Nc6 Classical line? Bd3 would have left the d4 Knight en prise.

I did not say that Nc6 is NEVER played in the Najdorf. I saif that it was an anomalous square for the Knight, and that Nd7 was the usual spot. 

Read my post.

Why couldn't 8. Bg5 be played in the game instead? It was played on move 10 also. The d4 knight was not en prise after 7. Nde2.

If 8. Bg5 Be7 why not 9. Bxf6?

This gets away from the topic, but why is it an inferior version of a Shvesnikov or Lasker/Pelikan?

Shvesnikov:

In the game continuation, if White plays 8. Bg5 Be7 9. Bxf6 then he gives up his better Bishop but Black can reply with 9. ... Bxf6 (instead of being forced to take with the g-Pawn, as in the Shveshnikov) and can later activate that Bishop with Bg5.

You mean inferior for black? I thought we were talking about white's play. This seems like e5 is the determining factor then, not Nc6 because that weakens the d6 square.

The main line Shveshnikov that I posted in #28 is slightly better for White. Your suggestion of 8. Bg5 Be7 9. Bxf6 in the game position (instead of the 8. 0-0 actually played) is better for Black... for the reasons that I just offered.

The e5 move (by Black) is not limited to the Shvesnikov / Lasker / Pelikan complex. It is also characteristic of some important variations of the Najdorf. In fact, the 6. ... e5 move was GM Najdorf's original idea behind 5. ... a6.

I see that MCO-15 is confirming much of what you are saying. I see e6 is also mentioned in addition to e5 as an option.

So, is there a site like chessables, but instead of clicking through a sequence of moves you are presented with a position and have to determine the opening position? If in fact all we have to do is look at a position and give it an opening name then I have gone about this the wrong way for years. Instead, I should be learning positions not opening sequences.

I'm not aware of any such site. It seems that Openings are always taught as sequences of moves, even though the IMPORTANT thing is the positions that are reached.

Quite seriously, one chess book that changed the whole way I look at the game was "Pawn Power in Chess" by Hans Kmoch. 

Pawn Power In Chess by Hans Kmoch - Chess.com

It was written back in the Stone Age (at least 30 years ago) but it was entirely focused on understanding the POSITIONS, rather than the move sequences, that characterize each family of openings. My rating went up nearly 300 points during the year or so that I spent reading, re-reading and re-re-reading that book.

I just got a nap in, and I am taking a second look now with position in mind. In the game in the OP, the game went:

1. d4 Nf6 2. c4 e6 3. Nf3 d5 4. g3

We shouldn't call this a Catalan until Bg2 is played. Is this correct?

1. d4 Nf6 2. c4 e6 3. Nf3 d5 4. g3 dxc4 5. Bg2

That is what was played in the game. The computer now identifies it as a Catalan. 

The following is a common opening sequence to reach a Catalan.

1. d4 Nf6 2. c4 e6 3. g3

This wasn't played in the game however. This is a common problem I have with identifying positions. I postpone judgment until more moves are made. So, in the Carlsen Dubov game I would wait until move 4. g3 was played and then call it a Catalan. 

My thinking has been to call an opening by the "ingredients" that make up the opening. 

1. d4 Nf6 2. c4 e6 3. Nf3 d5 4. g3

1. d4 Nf6 2. c4 e6 3. g3

Due to my ignorance of the position apparently, the 3. Nf3 and d5 mean something to further postpone judgment. I don't understand why we have to wait until 5. Bg2 has to be played in order to call it a Catalan. Likewise, I don't understand why we can't also call it a Catalan even if other moves were thrown in like this line.

1. d4 Nf6 2. c4 e6 3. Nf3 d5 4. g3 dxc4 5. Nc3 Bb4 6. Bg2

Then the question arises, if we had to wait until Bg2 was played before, why is the computer saying it is a QGD on this site? On another site it is labeling it an Indian Anti-Nimzo-Indian.

What's wrong with calling it a Catalan to begin with from move 4? What has changed from 4-6 to exclude it from being a Catalan?

nc3 is normally not part of a catalan setup, as it makes dxc4 more effective because it will be harder to get back the pawn, plus white allows bb4 tranposing back to a nimzo with g3 which i think is called the kasparov romanishin line of the nimzo. i checked my engine and it classifies 1 d4 nf6 2 c4 e6 3 g3 as a catalan even though i think only 3..d5 or a later d5 would make it so. maybe because d5 is still a possibility the engine just accepts it as a catalan. however, if white plays d5 himself or black goes for something else other than 3..d5 then it should leave the catalan universe.

the label anti-nimzo is only for white's 3rd move 3 nf3 as it avoids the pin. maybe the site doesn't adjust the nomenclature as more moves are played. after 4 g3 it became a catalan but after 5 ..bb4 it transposed to a nimzo g3. in the pure catalan white doesn't play nc3 and just allows black to take on c4 if he likes.

also, i think if black ever plays d5 but doesn't take on c4 (at least not early) it will be classified as a qgd. my guess is that the catalan sort of belongs to both the indian defence universe and the queen's gambit declined universe, but i still believe that it only becomes a catalan if the pawns are on c4 d4 g3 and e6 d5, so i only consider the catalan to be part of the qgd and not the indian defences.