Thanks, that is a super interesting talk! He posits one mechanism for how the octopus might be able to colour-match while being colour-blind, which is the presence of distributed light sensing pigments in the skin - a molecule for this, "opsin", can be found in the skin of cuttlefish and squid (and maybe octopi too, he didn't say).
He says that although the pigment they have found for light-sensing is the same one they have found in the eye, and thus only indicates that the skin may be able to "see" a single colour (i.e. is also colour-blind), the research suggests that other pigments may exist, or that some other mechanism like refraction of nearby chromatophores (pigmented and/or light-reflecting cells) may allow those pigments to detect more than just a single colour.
When I was first trying to imagine how a colour-blind octopus could colour-match itself to its surroundings, one thought I had was how a deaf person can sense sound via felt vibrations. Based on the frequency of the vibration, pitch can be inferred. In a way, that is sort of similar to what he suggests is happening with the light-sensing pigments in the skin - the information the organism requires is fed to it via its skin rather than the organs typically used for this task, the eyes.
[1] http://www.youtube.com/watch?v=mtIiSXRv5no&t=33m23s