I'm sure that ventilation makes a *huge* difference. One of the toasted imacs I received had died after its owner tossed her coat onto it without thinking that it would be a bad idea. When she retrieved her coat an hour or so later, the imac was dead.

As equill said, my data must be taken with a grain of salt because stuff usually finds its way to me only after it's been declared hopeless. That all but guarantees that I'll see a very skewed sample. That said, the over-representation of bad flybacks might be interpreted as reflecting a vulnerable aspect of the imac's design, nonetheless.