sure, I know fast Fourier transform and reinforcement learning use dynamic programming. but those don't represent most people's daily work. checking http://oj.leetcode.com/problems/ , you will see that the problem selection is very biased and academic focused.

it's like recruiting warriors based on gymnastics skills. I think it makes more sense to look for street fighters.