I totally agree that knowing the history is important. I guess that's a major reason that used bikes' market value starts at half the new price and goes down from there-- the simple fact that you usually can't know what kind of treatment they've had. I'm an electronics engineer, not mechanical, and I'm pretty ignorant about metalurgy except that I'm very detail-oriented and hate making all but the most minor purchases without researching the thing to no end. (Salesmen hate me because there's no way to get me excited enough to make a quick decision to buy!)
It is my understanding that steel has a threshold of how far it is flexed, below which you can flex it any number of times and it will never fatigue. It makes sense then that if someone is not doing the kind of uphill sprints simulated in the lab tests (the kind that broke my nice steel frame in only 10,000 miles with frequent hard efforts in the hills), or even non-sprinting out-of-the-saddle climbing, and they keep it rust-free, the steel frame will never break.
It is also my understanding that aluminum has no such threshold, so even if an aluminum frame is never ridden hard, it will, with enough pedaling cycles, eventually crack. However, the lab tests show the aluminum to be a lot better than I had thought before I saw the test report. Of course ideally they would test at least 10 of every model of frame on the market, but that would be basically impossible, at least from an economic standpoint. However, I have seen aluminum frames develop big cracks over the course of a few weeks where they got a somewhat innocent-looking dent. I have not seen this happen with steel.
