I don't think the salt caused the hydrogen, it was the lack of cooling that did it. When zirconium gets above 2200F it will ignite in a steam atmosphere and generate hydrogen. The mistake was the siting of the plant in a known tsunami zone and not adequately securing the diesels so they could not be swept away. In the USA you can't build a nuclear structure on sand.

As far as warning people, it is a double edged sword. Evacuating people, especially panicked people, is very dangerous. Sheltering is almost always a better option. Passing out KI pills 4 or 5 days after the reactors shut down is also reckless. KI blocks the uptake of radio active iodine by saturating the thyroid gland. Radio active iodine decays away to nothing by that time and scared people may overdose on it.

The problem with communication is almost always one of not knowing all the facts. When a disaster like this strikes, the ability to communicate to the off site organizations is severely limited. Even the ones fighting the casualty don't have the big picture, and some things, like boiling off the fuel pool, isn't something they expect.

Fuel rods were never ejected from the spent fuel pool, or anywhere else. Where did you get that idea?

I do agree that spent fuel buildings need to be more robust. I'm sure that will be the next NRC mandate.

US reactors are not designed to vent to another building. They vent to atmosphere and also have hydrogen re combiners inside the containment. Combustible levels of hydrogen can't form.