Possibly not applicable here but in the state of Australia where I live engineers require the plate for belts to be at least 3/16 thick by minimum 2"x 2", preferably 2"x3" with the edges of the plate where it contacts the floor to bevelled off so there's no sharp edge to act like a guillotine and slice through the floor. Also the nut for the seatbelt plate has to be welded, you can't tap a thread into the plate.
I watched that when it first came out. I hope he recovered. It looks enough like my car , that I won't ever take mine to a dragstrip.
I learned the hard way at a young age that when the *** end of a car tries to p*** the frond end, getting out of the throttle makes a bad situation worse. All the steering inputs in the world won't get you out of an over steer situation, you need to use smooth throttle input to re establish traction at the rear end. It's typical that you need to drive the car out of the slide, power on, but not at the same rate that caused the oversteer situation. So the typical teach is, counter steer, use the throttle to reduce power but don't get all the way out of it. If you get out of it, car weight transfers to the front, unweights the rear and increases the amount of oversteer. Power out of the slide. Looking at the video it seems the driver is late on his counter steering. At 1:09 the car is significantly out of sorts but the front wheels are still pointing straight to the car rather than down track. I could be wrong, but late, over corrections on the steering inputs could be a significant part of the cause of this crash.