The altitude hypothesis doesn't hold water.
If the window vacuum seal is intact, by definition there is no air inside to expand with altitude (or heat).
If the window vacuum seal failed, the air inside the window will expand with altitude (or heat) and exhaust where the leak is (because it is not sealed like a snack bag). The pressure inside will equal the pressure outside.
If for some reason you made a multi-paned window with air inside and then sealed it, the inside pressure would exceed the outside pressure with altitude and/or heat, but the difference of a few psi would not matter. After all, the GT window has a constant difference of 14.7 psi at sea level.
This is why I quit engineering school. I'm already OCD.
Sent from my SM-G950U using Tapatalk