If it's steel, and you use Rivnuts or similar threaded fasteners, you've improved your chances of success. Not having threads cut directly into the frame reduces the stress concentration considerably, though the enlarged hole offsets that to some degree. I'm no expert on the stress distribution in a bicycle frame, but I'd imagine the seat tube is a better place to experiment than the down tube. Two things to watch out for when using Rivnuts: choose the same material for the Rivnut as the frame to avoid galvanic corrosion; and avoid drilling the hole oversize to avoid having the Rivnut work loose and spin in the hole.

EDIT: I watched the video. The use of JB Weld Epoxy on the Rivnut is probably good insurance to keep moisture out of the frame and reduce the risk of the insert working loose. Thanks for posting that.