I'd have agreed on the extended lug design theory, until I saw 009.jpg. If that's two pieces of metal butted and badly welded, the root of what appears to be the crack is the end of a shoddy weldbead, then why is the joint/crack not even slightly perpendicular to the centreline of the tube? But then, what the hell leaves a crack initiator like the blow of a morning star, in a lug?