Observers Paradox.
Obviously the chance of getting the same hand at two tables is the chance of getting that hand squared ((1/52)*(1/51))^2, then cubed, etc....
The problem lies ( i think) in your tracking. While you may have XY hand at one table, 3 hands get dealt at another table, increasing the chances of it happeing. You would need to deal each hand at the same time to get a good sample here.
Of course, it could all be rigged.
Maybe it is happening, but it won't (shouldn't) continue.