So total momentum is conserved then, just the x component changes? And to change the wavelength the total momentum has to change? Is it not misleading when textbooks say "lasers produce photons in a momentum eigenstate because they have a precise wavelength" - surely when it diffracts and its momentum spreads out it is no longer in a momentum eigenstate/definite wavelength but you said it will be because total momentum/wavelength is conserved?Let me give you part of the answer. We are talking here about momentum in the transverse direction, not total momentum. For an EM wave, diffraction is about a change in direction, not about a change in wavelength or energy.