The decision for convection cooling was done mainly by fiat, not as the result of any careful engineering analysis.
Making excutive decisions is critical for the success of any organization. Hopefully they are only done after careful consideration of all of the available facts. But the reality is, production schedules often do not allow the luxury of careful research. Often a research cycle ends this way: the deadline has been reached before the completion of the research. The decision must be made on what information is available, complete or not.
You are talking in generalities, and then ascribing them to this specific instance; you are making the implicit assumption that this rational process was followed here. Let's look at the history, and judge for yourself: The engineers (mainly George Crow, chief designer of the analog board) knew that it ran too hot without a fan, so one was designed in. Jobs didn't like the noise, so the fan was taken out. He placed a premium on silence, and was willing to sacrifice longevity to achieve it; heat wasn't HIS problem. No deadline pressure forced a decision based on incomplete information. All involved knew that reliability would suffer. It is relevant in this context to note that, even as field reports of heat-related failures accumulated, the fan option was never exercised. From the 128's debut in 1984, to the retirement of the Plus in 1990, no fan was allowed in.
Fans crept into Mac designs after Jobs left. They were forced out again (of the imac) when Jobs returned, again to the detriment of longevity, and again over the objections of engineers.
So you see why I said that the design was by fiat. It was not due to any naiveté on the part of the designers. They knew precisely what would, and did, happen.