suffice to say the pax I met weren’t subjected to anything like 6 G and the seats pulled out of the floor.
The typical problem associated with seat separation with static design seats relates to seat and/or aircraft structure deformation which leads to the leg buttons being loaded asymmetrically, failing and away it goes. Hence the better situation with static seats bolted to structure in one way or another.
Some may recall that delightful ANR 40 chap, Rudy Paspa. One of his efforts involved designing and patenting (as I recall) an ellipsoid button which rotated on seat installation into the seat track so that the restraint capability was improved.
It was this deformation problem in mishaps which led to the requirement with newer dynamic seat certifications that the sled test required a preset misalignment of the seat to floor attachment to provide some comfort.
BCAR dated 1947 with a technical letter
Thank you, good sir .. saves me some trouble in researching the matter.