When I used to be the tame physicist accompanying bits of satellite from the US or Europe to Baikonur, as able to speak the language at both ends, there used to be endless discussions as to whether to encapsulate a printed circuit or not. Encapsulating adds grammes and the weight budget of a satellite is carefully calculated to the last gramme. Satellites are a whole new world of technology, with radiation-hardened components, thermal budgets worked out to the last Joule, and mechanical engineers putting in a lot of skull sweat into the lightest mechanical structure that will actually support the system.
The number of g's a payload can withstand and the g's the launcher is capable of are carefully matched. Ideally a launcher, for maximum efficiency, wants to accelerate as quickly as possible, this is only done with missiles. Launchers for commercial launches go up much slower to protect the payload, which before launch has been tested for acceleration, temperature and pressure.
Getting the bits of satellite from A to B is also taken into account. There's no risk of them being dropped 2.5m onto concrete by a clumsy baggage handler. Many satellites contain toxins (beryllium oxide), maybe radioactives, and sometimes explosive bolts. All of these have to be signed off by the captain. They are then loaded onto the cargo hold by the loadmaster, who treats the components as if they were the finest porcelain.
If even a single component is rated to only -30 C and at Baikonur it's -35, the launch is held. There's no point in launching a satellite that might fail.