one of the speakers being 180 degrees out of phase flips the positions of maxima and minima compared to what you'd get if they were in phase.
so I guess the 'helpful' answer is telling you to pretend the sources are in phase and then find the lowest frequency that'd give you a maxima at the given position...
or another way of looking at it is the microphone is a whole number of wavelengths distant from each speaker and the 180 degree phase shift of the sources is causing the destructive interference that gives the minima.
gcd (18, 24) = 6 so lambda is 6m
336/6= 56Hz