It's to protect from falling debris from the ceiling. How much it helps is debatable but it's best they have there in school. More effective on traditional bombing than nukes
In Finland we have bomb shelters everywhere, it's arguably more effective
It's also to give people something to do. Something to practice and focus on getting right. Gives hope and keeps people from getting caught in a panic loop.
In WW1 armies learned helmets were a good idea when artillery kicked up big chunks of debris killing unlucky soldiers when it rained down on them. Ballistic protection was an afterthought that came along later.
So yeah better than nothing I guess, same with tornado drills our schools have sometimes
They almost stopped using helmets again, too. The number of head injuries skyrocketed. Thankfully, someone pointed out to command that the helmets weren't causing the injuries, but converting fatalities into injuries. They hadn't been recording head injuries on corpses.
It also helps against what tends to be modeled and seen as the largest cause of injury during a nuclear scale explosion like that seen in Beirut, namely shards of glass, though it definitely helps survive falling beams in timber framed buildings.
Remember, thanks to the wonders of the inverse square law you are statistically far more likely to be in the area that gets light to moderate blast damage from the pressure wave rather than core of the blast.
Yeah, a nuclear blast is gonna be totally deadly within a particular radius, no matter what you do. And then at some larger radius, everything outside that radius will be safe, regardless of what you do. So the area in between is going to be the area where the response can make a difference.
And as you mention, the area of the "can actually make a difference" zone is much larger than the "dead-no-matter-what" zone, because it scales by the square of the distance. So if the outer safe radius is twice the inner death radius, the area of the in between zone is gonna be about 3 times the size of the death zone (π(2r)^2 - πr^2 = 3πr). If it's 3 times the radius, it'll be 8 times the area.
People only think about extremes. Why do we need seatbelts on a plane? Well, not for the crash, for the tons of turbulence you don't think about because you're wearing your seatbelt when they happen.
Just keeping people away from the windows could potentially prevent hundreds of thousands of injuries from burns and flying glass in the survivable area of the blast radius. It'd be really hard to overstate what a massive difference that could make when it comes to allocating medical resources in the aftermath.
We had an earthquake drill at one place I worked, where there was an outside door less than 20 feet from our desks. Another guy and I agreed, if there was ever an earthquake we were heading out that door and would wait in the middle of the parking lot until it was over and then help dig bodies out of the rubble. Somebody said, "What about falling debris?" It was a freaking 2-story building, but yeah we conceded we'd be taking a risk for a second or two as we sprinted clear - vs expecting our cubicle furniture to keep a collapsing building off us lol.
Having lived through a major earthquake - if it's a brick or concrete tilt-slab building, you are way better off inside the building. The risk isn't so much some random piece of something falling off, it's the entire facade of the building coming down on your head.
Correct me if I'm wrong, but if you're able to perceive the nuclear explosion and not even go blind, then you aren't close enough for your house to disintegrate like that.
The blindness thing is really only for a split second while the fission/fusion is actually happening. By the time the mushroom cloud has formed, the actual explosion was like 30 seconds ago.
If you see a full mushroom cloud, that means the glass in front of you is probably going to rapidly accelerate into your skull when the shockwave hits you.