I second this suggestion. Also, depending on how much gameplay experimentation and tuning you will end up doing, this could easily go long.
Many times, game designers believe very strongly in a certain approach, only to find very late that the approach won’t work, and it’s back to the drawing board, which may significantly extend the schedule. This happens less often for art, but when it does, it’ll be a complete do-over. So avoid this 
The sooner you can prototype any bit that you aren’t 100% sure will already work, the better for everyone. Gray box levels, stick figures, programmers saying “boom” into the microphone, are all totally legitimate things to play with when figuring out the basic game mechanics.
Character animations, effects, transitions, an environment modeling should come after all of the gameplay is already built, and everybody agrees it’s as fun as it can be.