The Mediterranean is an average of 1.5 kilometers deep, not 500 ft = 0.15 kilometers deep ; let's assume that is a typo in your source. But let's use your assumptions of a water channel 7.7mi = 12.4km wide and 1/8mi = 200 m high. We don't know the actual conditions at this future time ( assumed 12K years in the future?) so the blockage could be geological or man-made. An example of a geological blockage would be a Really Big Earthquake that dropped Gibraltar and some of the Rif mountains into the strait - the Bonneville Slide creating the Bridge of the Gods
900 years ago is a precedent. So are the ice dams that created the the Missoula Floods
. A volcano could also fill the gap with boulders and sand (most mountains in the northwest, like the small mountain I'm on right now, are giant sand piles, not solidified lava).
Meanwhile, the Niagara river is less than a meter high when it goes over the falls, and the combination of the three falls is about 1.1 km wide. The river feeding the falls runs downhill a meter or two over about 40km from Lake Erie - the speed limit is set by viscosity for this shallow river. If the water was 200 times deeper, the currents at the top would be 200 times faster, and the total flow would be 40,000 times greater. 440,000 times greater with a gap 11 times wider. Of course, the Niagara doesn't do that, because the Great Lakes Basin doesn't produce that much water. The Atlantic has a wee bit more water available.
The Gibraltar channel from Atlantic to Pacific is about 30km long, and the falloff is steep and wide at both sides. The channel will slow the flow, but with a critically important difference - the Atlantic is infinite, compared to Lake Erie, and when the water gets moving towards the gap, it will build up kinetic energy that either stays kinetic energy or turns into potential energy (bulging up). The main frictional effects are from the vast jumble of house-sized boulders smashing along faster than bullet trains near the bottom. The Zanclean flood moved water at 300km per hour, through what is depicted in illustrations as a much longer and narrower channel. So the OTC flood will be at least that fast, possibly faster. The OTC flood might move 10 cubic kilometers of water per minute when the flow eventually gets up to speed, increasing in volume as it cuts a deeper and wider channel through Gibraltar. That is 90,000 times Niagara's 110,000 cubic meters per minute. The power generated by a 10 km3
/min OTC flood, dropping 1500 meters near the beginning, is 2.5 terawatts, a 2MT bomb per hour.
Regulating all this is inertia, not friction. Instantaneously remove a dam, and a wall of water does not flow immediately; the pressure difference accelerates the water, from slow to faster to FAST. The pressure difference gradually spreads out along the length the channel, the same amount of momentum change occurs, but it is spread out over a larger volume. The velocity keeps increasing until the pressure gradient is zero, and only starts slowing down when filled basin gets higher than the source basin, creating back pressure. For massive volumes like the Atlantic and the Mediterranean, the rampup may take weeks, and the stopping (after the basin is on average filled) may take weeks, too. Water 1000km out in the Atlantic is still moving towards Gibraltar after the Med is full, and will keep moving until back pressure (caused by water higher than mean sea level) in front of it slows it down again.
In the beginning, the time currently being portrayed in the OTC, the accumulation of water is square law for a fixed dimension Gibraltar gap, perhaps cube law if the gap is eroding wider and deeper. The water isn't spreading uniformly as you might expect, momentum sends it faster eastward than it veers north and south. Perhaps a few hours of advantage for Megball and their homies.
I doubt that GLR has done the hydrodynamics himself (he might have run some existing numerical code), but I'm guessing he's found the numerical simulations of the Zanclean Flood and is turning those abstract teraton water models into dramatic spectacle. Someone with more research or detective time might find out who he talked to or what research papers he might have read.
People die in tsunamis and flash floods partly because their inner "water flow calculator" is calibrated for small scale viscous flow, not the inertia movements of huge masses of water. So you have to do the calculations and think about conservation of energy and momentum, and the relatively narrow zones where the energy can slowly dissipate into the rock and the sky. I doubt Cuegan or their people will do these calculations.