I don't know why this issue is still being discussed... The culprit has been found via telemetry, it's up to iRacing to fix it...
And no, the video doesn't show 2FMSH... Everyone half knowledgeable agreed that locking the front tyres would cause them to wash out, thus decreasing the yaw angle... But for that to work the rears would have to roll... Now in your video, the BF1 didn't have enough brake force to lock the rear wheels and nKpro and rF2 lacked any perspective to see whether the rear wheels were locked too... I didn't see the rears lock up in the iRacing bit, but the pCARS part showed that the car stopped correcting itself as soon as the rears were locked and skidded along with the same amount of yaw angle...