Why is acceleration significant to generating power in racket sports? I'm reading a paper on badminton stroke power (see Figure 4 on page 8). It says:

For most of the trials the racket head reached peak speed just at the time of impact. The racket
  head showed great acceleration just before impact where the racket head speed went from around 10 m/s
  to its peak value [35 m/s] in less than 0.1 seconds.

In related text:

it seems reasonable that
  skilled players will produce higher angular velocities of the glenohumeral internal rotation, elbow pronation
  and wrist flexion in the execution of the forehand clear stroke compared to the less skilled players.

From a physics perspective, why does high acceleration / high angular velocity in a stroke produce more power compared to brute speed (without technique) of the same stroke? Isn't momentum = mass * velocity, so these two are just the same?
I don't know physics.
 A: Several things here.
First - how efficiently you transfer momentum depends in part on whether you hit the "sweet spot". The location of the sweet spot is a function of the instantaneous center of rotation - an almighty swing from the shoulder puts the center of rotation further back so the same velocity of the racket might be less efficient. That depends on the design of the racket - but assuming the racket was designed for a particular stroke, then it will be less efficient when you use a different stroke.
Second - in many racket sports (and particularly golf) the swing is that of a compound pendulum - initially the racket/club head lags the movement of the hand(s), but in the last part of the stroke it catches up. This transfers energy from the arm to the racket. The advantage is that the racket, which has further to travel, spends relatively less time at its top speed. This means much less total drag and thus more energy for hitting the shuttle.
In principle, if you have the exact same motion of the entire racket at the moment of impact, then the history doesn't matter. But the technique used gets the highest velocity at the lowest energy cost. And if you have to execute hundreds of strokes during a match, that really starts to matter.
A: 
For most of the trials the racket head reached peak speed just at the
  time of impact. The racket head showed great acceleration just before
  impact where the racket head speed went from around 10 m/s to its peak
  value [35 m/s] in less than 0.1 seconds

Probably tou have misinterpreted the text, it nowhere says that the acceleration gives more power, what is important is that the head reaches max speed before it hits the ball and, of course, not after

the longitudinal axis rotations reached the highest angular
  velocities of all joint movements, supporting the idea that such
  joint movements may play a crucial role in producing high racket
  head speeds.

The paper is just reporting the how and why that happens, is not saying that if the highest speed is reached earlier and the speed is alsready constant the power of the strike will be less powerful.
As you say the power depends on mass (according to the parameters reported in Benichiwa's post) and speed which make up momentum. I doesn't really matter if max speed is reached 1/100, 1/10, or 3/10, etc., before the impact
A: They are the same thing. Parts of the racket though produce more force than others, suppose I completely missed the strings and hit the metal part (that's why without technique is so important). Usually, when you play you hit in what is called the sweet spot which is a vibration node and although it may not give you the biggest pop or bounce it gives you more speed, because it is closer to the top of the racket. Here is an article about the physics of tennis rackets http://www.physics.usyd.edu.au/~cross/tennis.html.
