Dyson seems to be making a more abstract argument — that if the domains where a graviton could be detected are inaccessible to us, then they effectively do not

Dyson seems to be making a more abstract argument — that if the domains where a graviton could be detected are inaccessible to us, then they effectively do not exist (if a tree falls in a forest...). At least that’s what I took away from the description in the article and his dismissal of the proposed experiment as not having a bearing on his objections.