The Meran Defense,one of the most complex queen pawn openings, was also entirely Rubinstein's idea, but it was named after a place rather than a person. And of course, all the great players have played this complicated opening (Kasparov, Karpov, Kramnik, Anand).

Developments in the opening are a bit of a tempi struggle and related to the orthodox defense: for years, Black's c6 was considered a wasted tempo since c5 would ultimately need to be played. Capablanca (and Showalter) demonstrated that the tempo lost by c6 was offset by White's extra tempi used moving the king's bishop. Ultimately, the source of White's advantage is Black's "problem child" bishop: the Meran would seem to confound that problem for Black, with the hedgerow of pawns on white squares, but he's often able to free himself and equalize the game.

8.Bd3 a6 9. O-O c5! which was Rubinstein's idea to transpose into a solid defensive variation of the Queen's Gambit Accepted, with the only difference that both Black and White have used an extra tempo to reach the position.

There's a case for attributing it to Osip Bernstein:

Here's a Chess Note by Edward Lasker dealing with the Meran Defense (mainly the false assumption that Tartakover popularized it at Meran 1924 also) http://www.chesshistory.com/winter/...

What about 28. Bxe5 as a move!!??

This is indeed the game generally recognized as the first Meran Defense [Counterattack].

Grünfeld’s <9. 0-0> is plausible (and the second most-often played choice in the CG database, with 125 occurrences at this time), but it is weak. (Black scores 56%, specifically: White wins: 28%; drawn: 32%; Black wins: 40%.) There is just too much happening in the center for it to be advisable for White to castle here.

Grünfeld did at least play <10. a4>, which gets back to a position with a plus score for White amongst games currently in the CG database. It is nevertheless better for White to play <a4> on move <9>, since the pressure against the b5-pawn then forbids <9. … c5>.