I think that this approach should work rather well (I did very simillar project few years ago). Things you might consider:

 1. You probably might want to use pre-emphasis filter on a signal.
 2. In addition to MFCC features you can also include $\Delta$ and $\Delta\Delta$. Some theory can be found for example here: [click!][1]
 3. Comparing against more templates in my case largely improved recognition rate. I think that the easiest way do so is by using k-NN algorithm on distances returned from DWT.

Good luck!

  [1]: http://speechlab.eece.mu.edu/papers/Ye_thesis.pdf