Dave: I talked to a weldor about this, and they said the issue here with safety gear is that you weld with lower-grade filler to reduce the heat affected zone. They don't (contrary to what I expected) put airframes or car cages into an over to stress relieve them, so they need as small a HAZ as possible. Diandra was given some specialized information on an exceptional process and generalized from it; probably the reason it was emphasized in the forum she was in was due to the exceptional nature of that type of welding. Diandra, if you're still reading along, thanks for writing here, there, and everywhere.

I asked over here and, while the responses somewhat support my position, they don't link to any documentation. I did find AWS has a set of standards which talk in depth on the issue, but at a couple hundred bucks a pop, I'll pass.

A couple standard tests for weld quality are a tension test, where a section of weldment is pulled apart; and the bend test, where a section of weldment is bent on the weld. If it fails on the weld, the weld (and if it is in a class, the weldor) is failed and does not pass go. These guys have pictures of both fixtures on site. All of this is a long-winded way of saying a proper weld is stronger than the base metal. Obviously, here, we're dealing with a defective weld, so you're quite correct to say "When the piece hardens, you have something that, while not as strong a a solid piece, has enough strength to resist whatever pressures might be pushing on it." However, in the general case, the weld will be stronger than the base metal. Loved "The Physics of NASCAR".

@Swaroop, BlackSheep does nothing interesting; it is, in fact, a DDoS against facebook.