This answer is a counterpoint to Prucha's principles, as given in xaisoft's answer, in particular the first principle:
The Albin is playable only if it is treated as a pure gambit. Black should not try to regain the pawn by attacking White's e5-pawn, but must continue in gambit style with ...f7-f6
.
As mentioned in the question, the Albin saw a revival at the highest levels in the last few years, primarily in the hands of Alexander Morozevich. And the approach that he has taken is precisely that against which Prucha recommends; namely, Morozevich (and Nakamura too) have responded to the mainline fianchetto with 5. g3 Nge7
, with the intention of following up by ... Ng6
in order to gang up on the e5 pawn. According to Nigel Davies' Gambiteer II:
This has been an Albin backwater for years, but now it looks like it might be the best way of playing the gambit. Naturally White can try to hold on to his pawn, but this can often prove to be a dangerous course.
Prucha's third principle, about maintaining the d4-pawn, seems to be generally good advice (though I wouldn't take that "at any cost" too literally). A nice example game where Black's play revolves squarely around the d4 point is Sokolov - Morozevich (Corus 2005), in which Morozevich maintains the pawn on d4 up until the point when it became a strong passer with 27. ... d3
, and Sokolov resigned after move 34. In general, though, the Albin seems to be an opening where calculation and concrete details in particular positions carry the day more than a handful of guiding principles will.