Why does 7.g3 score so high for White in the Blumenfeld counter-gambit accepted
The thing is that all the other responses seem to be more or less equal for Black.
So the approach the players you mention take is like a Grunfeld - pinning the d-pawn then attacking it.

The pawn on g3 prevents cheapos on h2. Without cheapos on h2 Black's strategy loses a lot of it's bite.

Looking at games that have been played so far in 2017.
74 have played 7.e3.
0 have played 7.g3.
Generally what this means is that someone found something for black after 7.g3. That would be my guess.
This is actually a great illustration of why you shouldnt play a move simply based on %.

Looking at games that have been played so far in 2017.
74 have played 7.e3.
0 have played 7.g3.
Generally what this means is that someone found something for black after 7.g3. That would be my guess.
This is actually a great illustration of why you shouldnt play a move simply based on %.
Pretty much. They're trying to avoid this game:
Another idea is 7.Nc3 with the idea that White wants to play e4 and wreck Black's pawns no matter what. After 7...Nbd7 8.e4 d4 9.e5 it gets kinda crazy.

I seriously doubt that the result of a 1922 game has suddenly caused everyone to abandon the line in 2017.