The 20ms in Neuroscience?

Last week in Steve's seminar, when proposing the non-Hebbian ISDP, he said 20ms seems to be special for the paring of Perforant pathway and Schaffer Colateral Pathway. It's kind of strange that 20ms to be a special number, as I can tell from his own feeling.

Today, I'm reading this 2003 nature paper about hippocampal cell assembly. They propose 25ms to be the best timescale for sychrony. And they think this to be physiologically significant in 3 ways: 1. it matches the membrane time constant for pyramidal neurons in hippocampus. 2. it matches the period of the hippocampal gamma oscillation. 3. it matches the effective window for synaptic plasticity.

So, 20-25 ms, the lucky number in neuroscience?