Yes, positive reinforcement works on essentially all pokémon. All people as well. Basically, anything that can comprehend a rewards-based system can be trained using one.
Just think of training a humanoid pokémon like training a child. Sure, you can simply teach a child how to do basic things such as read or tie their shoes or whatnot, but they become far more engaged if the learning was fun. This is why teachers in certain school systems reward their children for attempting to learn, either with literal rewards or by making the learning in itself a wonder. By contrast, try forcing a child to learn something they clearly have no interest in using methods that in no way makes the learning itself a pleasure. Chances are good that the child will refuse, sometimes rather vocally.
Humanoid pokémon are the same way, as are any other pokémon, and that’s the core principle behind positive reinforcement. The main philosophy is making the training seem worthwhile to the subject and helping them equate the task with goodness by associating the feeling of being rewarded with the task itself. If you want a hitmonchan to punch a target quickly, you make a game out of punching things or add a reward for doing so or both. Eventually, the hitmonchan understands that the faster and more precise it punches, the happier you and he will be, regardless of whether or not there is an actual reward for doing so.
Honestly, this method is the most humane. Simply teaching a hitmonchan the physics of punching or why he needs to punch won’t be as effective at helping him remember what he learned or why he needs to punch, and using negative reinforcement (threats and abuse) not only does more literal harm than good but also is both fantastically illegal and the sort of thing that would make you a terrible person.
The Pokédex According to Bill 