Answer for question 5 (c).

Let b=2^k.

Dynamic programming takes O(n*2^k) time.
Recursive algorithm takes O(k*2^n) time.

Those two complexities are very symmetric over k and n.
For small k (or b), (1) runs fast.
For small n, (2) runs fast.
 
For large k, (1) runs out of space.
For large n, (2) does not run out of space.