Not to 'stir the pot' but there's a distinction here that you're missing. In that video, as in any case where there is an actual lead spray, the water had to get under the surface of the melt before it flashed to steam. Standing a few feet away and giving a mighty squeeze to a turkey baster can give the water stream enough inertia that it will penetrate the melt surface when it hits. The same thing can happen when a raindrop has 2000 feet to accelerate before hitting the surface of the melt. If a fellow could 'dribble' enough he actually could safely pee in the pot...except for the cloud of rising steam that might cook his weenie
If the video guy REALLY wanted to demonstrate a tinsel fairy he should have tied a saturated cotton ball to a wheelweight and tossed that into the pot!
Gently pouring the water on top is an entirely different thing than forcefully shoving it under the surface. As long as ALL the water remains above the melt then the steam has someplace to go when it expands. The problem is that it's not all that hard to accidentally drag a bit of water below the surface..especially if it's hiding on something you're adding to the pot.
I have some 1lb ingots that have been stored in the garage for two years now, but they once got wet and so they have a little layer of oxide on them. If I don't rest them on the edge of the pot for a few minutes to let the tiny bit of condensation inside the oxide layer bake off I'll get a small eruption when I add them. Again, it's because the water is being carried under the surface.
So, to summarize:
1) Water ONLY on the surface won't cause a tinsel explosion
2) Water below the surface will cause one
3) The threshold between (1) and (2) is an easy one to cross accidentally
4) Keeping water away from the melt is, in general a good idea.