[kurthr]

Even those MEMS devices (like sealed poly and Bosch etched bulk resonators) don't use a full vacuum. It's inconvenient and difficult to control. Instead a controlled amount of non-reactive gas is included to give repeatable results (eg damping). Having a known dielectric even close to 1 is fine and atmospheric pressure effects are negligible (eg <0.1%).

In the case of electrical interconnects I'd be surprised, if they weren't sealed with an atmospheric pressure inert gas (N2, CO2, Ar) just to keep moisture out or at least drying agent. Moisture (and dog forbid, condensation) is more likely to cause catastrophic effects than dry air. You could easily go from 1.0005 to 10... and that's ignoring conductivity and dielectric loss.