Well, aside from the mention of there being only 100 million neurons in the brain (there are actually 86-100 billion of them), the general idea that he communicates is correct. The activity pattern of a neuron through time (i.e. temporal activation) is how cognitions are encoded. NeuronA firing 5 milliseconds before NeuronB is a VERY different pattern than NeuronA firing 5 milliseconds after NeuronB. We see this very easily in the low-level retinal bipolar cells.
This is seen most readily in so-called "center/surround" cells. Although the idea of center/surround is simplified to the point that it ignores some of the more complex aspects of the phenomenon (even by renowned scientists), it still teaches us an interesting thing about how brains work.
It has been shown that certain Retinal Ganglion Cells possess very idealized center/surround properties. This means that some of them react to light stimuli in a way that causes rapid action potentials when the light is shone only in the center of the RGC's receptive field. Other RGCs demonstrate a reaction of rapid action potentials only when light is shone in the surround field of the RGC's receptive field, with no light in the center. This causes both cells to have inverse properties, where one cell is active when the other one is not. Then after a small delay, the activity of each cell flip-flops so that the one that was active is now inhibited and the other cell becomes active.
Other RGCs also react very complexly to strange patterns of light stimuli. After studying these phenomena for a while, it becomes clear that the timing (onset or delay) of the action potentials from a neuron communicates very important codes to the postsynaptic neurons. So yes, a huge part for neurons to interpret information from a string of action potentials is largely rooted in the temporal code, relative to other neurons.
