According to the chess data base, This was played ~300 times, with an average win for white as 42.2%, draw 19.8%, and black 38%, if white plays Nc3 after that.+when.board.is.flipped.+
King's Gambit declined

I don't know. I think the independent point of this move is that after 3.Nf3 Black can play 3...f5 which leads to complications. 3.Nc3 is a Vienna gambit which White may not want to play.

I don't know. I think the independent point of this move is that after 3.Nf3 Black can play 3...f5 which leads to complications. 3.Nc3 is a Vienna gambit which White may not want to play.
Quite right- 2...Nc6 is usually associated with the ...f5 sortie.
This is just another sound way (out of many) to combat the king's gambit.

But why opt for the same weaknesses as white? ...f5 is usually associated with the Latvian. Unless white plays 1.d4 or some Nimzo-Larsen and English lines an early ...f5 should be avoided IMHO.
Speaking of early ...f5 moves that actually work the Leningrad is one of the best options against weaker players (though 1...Nf6 and 1...d5 remain objectively better, perhaps 1...Nf6 is more accurate than 1...d5 as it's more flexible since a knight ends up on f6 anyway) since you soundly "violate" so many principles, such as ...a5 Na6-Nc5 and ...Qe8 so weaker players think to themselves "there's no way in hell black's going to win" and push themselves too much and usually play a gung-ho Ng5-Ne6 thinking they're making progress when all they're doing is plugging up the e6 hole with their pawn, which actually helps makes black's king safer.
If a player who is black declined the King's Gambit by 2...Nc6,
Then it is advantage for black or white????