It was known for a long time that it sufficed to prove it for $n \ge 2d$ --- indeed, if $n < 2d$ then every pair of vertices shares a common facet, and iterating if necessary, one eventually ends up with a case where $N = 2D$. (See Ziegler's book Lectures on Polytopes, p. 84).
Now it is known that the Hirsch conjecture fails in general, but does it hold for $n < 2d$? So far I haven't been able to think of a simple proof, and actually I don't even know if it's true. The counterexample due to Santos has $d=43$ and $n=86$.