In addition to the energy considerations of a human generated sound wave, there is this to consider too:
As you state, sound is a longitudinal wave of pressure variation through the air. Let's suppose we have the capacity to generate a VERY strong sound wave.
At the atmospheric level of the thundercloud, where water vapor is forming droplets, assuming the temperature stays relatively constant, the phase diagram of water would tell us that an increase in pressure will likely only increase the size of the water droplet (i.e. more water will be forced into the liquid portion of the phase diagram). Conversely, a rarefaction or decrease in pressure, will likely only decrease the size of water droplets.
Since compressions and rarefactions follow one another, the net effect of this on water droplet size would likely be zero.
(Adding in the fact that things won't stay isothermal further supports this. A pressure increase will lead to a temperature increase, which will work opposite that of the pressure increase: droplet size would likely decrease. So the net effect of a pressure increase on droplet size is likely net zero. Same for a pressure decrease when taking into account temperature changes due to the pressure change.)