Well the first database I found that had Nb5 after Bf5 showed white scoring 2.5-0.5 out of three games played. As this database also had amateur games as well, this might not be the most accurate judgement, and three games isn't enough to prove or disprove an opening anyways. I believe that your opening might be playable for white, but as it offers black easy equality and good chance for an advantage I wouldn't ever play it. Also, I don't think Bd3 giving up your better bishop for a knight looks that great. Try finding a safer way to defend the fork. In addition, your opening lacks kingside development, which might be a problem if black can begin to initiate an attack. I wouldn't play it, but I don't believe it is an immediate loss.
Note:
I have an online game in a similar Nb5 line going on currently (grunfeld, not qsfl or Whaite-Harrison - and I was playing black). I made solid developing moves, and at this point I should be winning. I will post the game when I finish it so we can analyze the opening and why exactly I gained an advantage. In addition, nobody has analyzed my Nb8!? that I gave earlier. The idea is that it transposes to a solid line in which black scores very well, and reroutes the knight from a6 to a better square, such as d7 or c6, rather than trying to attack with Nb4. I'll look back to the exact post and post the line again.
Actually, in the master's database on chess365.com, there is over 170,000 games with the first move d4, and NONE of them feature that position.