I would guess the latter as my father and this guy nicknamed "Blast King" did the calcs and building designs for structures around rocket test stands. Blast pressures were what they cared about.
They had a terrible time working with hydrogen. The stuff leaked like crazy and drove their mechanics nuts. One time, a mechanic waved his hand over a pipe to try to feel where a leak was coming from. He found it. The jet of gas pierced the skin of his arm and inflated it like a basketball.
Here's the other chart they used. It uses TNT blast as a reference. Figure C4 is ~four times as powerful as TNT per weight so that 100 cu ft of mixed H2 would give a blast equivalent to 15 to 16 pounds of TNT. You'll see from the graph that 15# of TNT gives a pressure of ~2.5 PSI at 50 feet.
Keep in mind this chart dates late fifties/early sixties. ....And I suck at math.
Looks like walls to 200' and windows to 500' are likely in trouble.
If the 100 cubic feet of hydrogen-air mixture were surrounded by a lightweight and fragile balloon, the peak pressure would be nowhere near that produced by TNT. Inside an automobile, pressures would get higher because of the mass of the windows (even after pressures exceed the breaking point of the glass, it would take a few moments for the glass to be pushed out of the way). Nonetheless, the failure of the containment vessel would limit the peak pressures reduced therein.