Notice that the "bridge" beneath the metronomes is oscillating in time with them. This is why soldiers break cadence when marching across a bridge.
I've heard about that, with soldiers; it was explained as having something to do with a frequency that mechanically resonated with the bridge and could make it fall. Does that make sense? I'm wondering in this case if some fuzzy initial majority tendency got the table moving in a way that influenced more and more of them? In other words, maybe it happened
because of a subtle movement of the table?
A modern, well maintained bridge wouldn't fall from marching over it, but old bridges built centuries ago could. Also, any bridge suffering from war damage could fall. (Imagine that -- soldiers marching across a bridge during a war.)
I've felt the High Falls pedestrian bridge vibrate form just one person jogging across it, and that bridge was built over a century ago.