This is interesting, and may have useful applications, but it's not a validation of quantum computing as it's usually conceived, in that it appears to be doing the computation with multiple photons, so it could be relying on different photons being in different states. True quantum computing relies on the Copenhagen interpretation of the current framework for Quantum Mechanics: that the probability distributions it operates on are, in some sense, real, rather than merely reflections of our uncertainty regarding the system's state. It assumes that by operating on a superposition of these probability distributions
within a single system, multiple computations can be performed simultaneously. In the system described, that apparent superposition in the probability distributions is potentially being "faked" by superposition in the concrete, individual states of the multiple photons involved in the computation.
But I'm biased. I'm quite dubious about the Copenhagen interpretation of Quantum Mechanics, and I think quantum computation is a reductio ad adsurdum of its conclusions, which may one day be used in its experimental disproof.