But you're broader point is true. <...Qc7> is common in the Kan and would seem to be a natural response to <6.Bg5> were it not for the tactics of <7.Ndb5>. I wonder if Dorsa played 6.Bg5 as a trap.

When both players are rated 2200+ (master level) ChessTempo has 84 games with 5...Ne7; White won 32 (38.1%), Black won 24 (28.4%), and 28 were drawn (33.5%). Still reasonably even considering White's average advantage of 35.0% wins compared with Black's 25.4% wins and 39.5% draws.

But after 6.Bg5 the situation changes, although there are only 9 games in the database so it's hardly a statistical sample. Of these 9 White won 5 (55.6%), Black won 2 (22.2%), and 2 (22.2%) were drawn. When both players are rated 2200+ the situation remains about the same; 7 games were played and of these White won 5 (57.1%), Black won 2 (28.6%), and only 1 (14.3%) was drawn. So clearly 6.Bg5 looks somewhat statistically superior to the 3 more popular responses to 5...Ne7:

(1) 6.Bd3: 66 games for all players, White winning 28.8% and 39 games when both players were rated 2200+, White winning 33.3%.

(2) 6.Be3: 45 games for all players, White winning 40.0% and 25 games when both players were rated 2200+, White winning 44.0%.

(3) 6.Be2: 42 games for all players, White winning 47.6% and 19 games when both players were rated 2200+, White winning 47.3%.

There were 10 games listed with 6.Bg5 with Black's most popular response being 6...Nbc6 although I can't see why; there were only 4 games listed and White won 3 of them, with the 4th one drawn. There were 4 games listed with 6...f6 7.Be3 Nbc6 (one of them involving a move transposition), White won 2 and Black won 2. So, in the absence of adequate data, clearly 6...f6 gives Black the best practical chances.

After 6.Bg5 Qc7 there were only 2 games listed, this one from 2013 and another one from 2010 that was drawn. But the 2010 game continued with 7.Bd3 and not 7.Ndb5 as this one, when White sacs a piece in exchange for a long-lasting initiative and control of the d6 square. This resulted in the loss of Black's queen for a knight and the resulting loss of the game. So I'm not surprised that 6...Qc7 was not repeated.

After 6.Bg5 f6 7.Be3 Nbd6 there were 3 games listed when both players were rated 2200+. White won 2 and Black won 1. So clearly 6...f6 gives Black the best practical chances. Opening Explorer lists two of them:

J Friedel vs K Mekhitarian, 2008 (0-1)

Kavalek vs A Tsvetkov, 1964 (1-0)

But even these two results might not be traced to an opening advantage. In the game that Black won he had a 13 point rating advantage (effectively equal playing strength) but in the game that White won he had a 240 point rating advantage. This translates to an expected 80.2% scoring percentage so White might have won in spite of the opening, not because of it.