The importance of coherence in phototherapy has been questioned over the last two decades, with the arguments largely being based on; 1) Lasers are just convenient machines that produce radiation, 2) It is the radiation that produces the photobiological and/or photophysical effects and therapeutic gains, not the machines, and 3) Radiation must be absorbed to produce a chemical or physical change, which results in a biological response.
Whilst these conclusions are, in essence, true, they neglect to account for the effects of laser speckle in vivo. In a proportion of individual laser speckles the intensity is higher than the surrounding environment, and the light is partially linearly polarized. This is important because the probability for a photon absorption event to occur largely depends on intensity and the photon absorption cross section of the molecule (which in turn is influenced by polarization and several other factors).
In superficial tissue, where the photon flux is high (less absorption has taken place), it is easy to reach necessary power density thresholds without the benefits of laser speckle. However, in deep tissue where the photon flux is extremely low, the increased probability of photon absorption from individual laser speckles increases the probability of reaching the necessary power density thresholds. Because of the non-coherent nature of radiation from light/IR emitting diodes speckle does not occur in the tissue with LED therapy, which may explain why head-to-head comparisons between lasers and LEDs in deep tissue seem to be in favor of lasers, and super-pulsed lasers in particular.