I wonder if somebody vaguely remembered the effects of helium fill, which actually does reduce sound transmission, and vaguely remembered that argon-filled windows are a commercial product, and conflated the two because they're both noble gases.
They didn't even start with museum-grade glass for optimal filming, which to a absolute layman like me seems like they just hadn't bumped most of these problems before.
Meaning there was no one in the world who really knew how to build such booths. Moreover, they discovered required features of the booths over the years. In such a scenario, the early mistakes are very understandable.
[1] Think Street Fighter, or FIFA
[2] I don't think games like Starcraft has such low time tolerances on information flow. Even cheers in Dota can be enough of a clue.