This is a bad idea to postpone Nf6 for two reasons :
a) White can try 3. Nc3 then Black can't take on d5 without creating a hole on d5 for the white knight. After 3. Nc3 if you play 3... Nf6 to exchange the N when it arrives on d5, then 4. e4 creates more problems as now 4... exd5 simply loses to 5. e5. The same happens after 3. c4 b5 4. cxb5 Nf6 5. Nc3 exd5 6. Nxd5 and Black runs into all kinds of problems (6... Bb7 7. e4!? and the like). And 4... a6 is a very poor Benkö-like as in the Benkö the bishop is really needed on g7 to pressure white's queenside.
These positions with a hole on d5 are fine if Black has enough pieces development to compensate for it, but it's not the case here.
b) after 3. c4 b5 White can simply play 4. e4 with quite a large advantage.
For all these reasons and more it has been established that the inclusion of c4 and Nf6 at move 2 is a general improvement for Black.
I have had a lot of recent success as black using the strange opening above, especially since it can lead to winning a piece with two different traps (variants in the board above). As far as I can tell, there has not been many games with it. Presumably, the opening is suspect, but stockfish rates this position a +0.30 which I feel is reasonable for black.
I know it starts off as Old Benoni, but basically, I skip NF6, in order to keep the option of playing Qf6 available. It plays a bit like a benko gambit, but to the best of my knowledge, it isn't actually a benko without Nf6.
How would you evaluate this opening, and why do people not play it more?