I am preaching a simple philosophy when it comes to heat in
microelectronic systems: When possible implement reasonable solutions
to lower the heat stress. You attitude is "Go ahead - run your system
at the bleeding edge."
I disagree fundamentally, even though I would never be able to prove
it to you.
As far as experience goes, you are a technician who has seen a lot of
broken computers. But you are not a scientist. I am. End of discussion
about credentials.