As a follow up to my previous post, here are some statistics to chew on:
The Chess Assistant HugeBase has 2,675,000 games in it, most of them played by players rated from the high teens to GM. Of all those games, 1e4 c5 was played in 549,459 games. In those games, Black averaged 49 points out of a hundred games played (each win gives a point and each draw gives a half point). The Sicillian is clearly a very statistically successful defense, but very common.
From the same data base, 1e4 d6 was played in only 60,493 games. Very uncommon. Black averaged 45 points from every hundred games played. So the Pirc is statistically less successful than the Sicillian. But since it is so less common, not as many people will know how to attack it in the best fashion, and playing it against people who aren't masters or above may well help you gain an early advantage, particularly if you personally know the Pirc well enough.
The thing with the f3, Be3 and Qd2 setup for white in the pirc is that black should never castle kingside too soon. He should wait a little because if he castles too soon he will be under very heavy attack. Often there is also a possibility to castle queenside for black in some positions.
If white should castle queenside then black attacks from the queen side but how would tackle white if he castles kingside?