I was playing a friendly match and suddenly this popped out at me. 

As you can see Black has a rather passive position and I played Nxe5, whereapon he slipped up with dxc3? and I played Nxf7, winning material.

But what I want to ask is:is the idea of 2...Na6, with moves like Nc7, covering b5 from the knight and allowing the pawn advance d5, valid?

I think it's a fairly strong novelty that can be used as a surprise weapon(I haven't peered into any tactics or analysis as yet, so there may well be a line that wins a free pawn for White or something; I'm assuming there's no such line) in fast-paced games. 

I used to play 2...Na6 with the idea ...Nc5 to pressure the e4 pawn.

I personally think it's a terrible idea to play Na6-c7, it's simply to slow. However I think black should do it with the main intention of playing e7-e5 rather than d7-d5.

Oddly enough I've played against this before in a blitz game a few years ago. Which is really strange since I probably don't even play 1. e4 1% of my games. I consider myself quite good at handling unorthodox moves like this, but this game I hardly impressed.

Conzipe, what about supporting a b7-b5 pawn push to attack the queenside? 

The main defect of 2 ...Na6, besides putting a Knight out of play and committing to a waste of several moves in the opening to bring him to any useful square, as I see it, is that after 3 d4 cxd4 White may play 4 Qxd4 with impunity.  Moves other than 3 ...cxd4 are no better for Black.

2 ...Na6 is as weak as it looks, and that's rather weak.

Lazarise the guy with 6.e5.