Not necessarily. Evolution does not mean improvement, just adaptation. Unless there's some sort of biological pressure, such as a predator or an environmental risk, change won't happen or will happen much more slowly.

The time scale is also important. In very cold environments, such as the ones described, everything happens at a much slower pace. Consider it this way: heat is a proxy for energy (technically it is energy, but it's not all the energy). Energy is required for change. Very cold environments mean less energy, which means less change over the same time period.

That's ridiculous. There is no law that says evolution can optimize itself out of any constraint. And even if it does, it might not look like much. For example there are tons of bacteria in the sky just floating in the wind, but unless you look closely you'd pretty much never know. And for most of Earth's history you wouldn't have seen multicellular life. It took a long time for the giant forests you see to evolve. And even then, you don't have that kind of massive biomass in places like Antarctica or deep underground.

Evolution is a pretty tricky thing. If some improvement can't easily be reached through iterative random improvements, then it won't be. And it easily gets stuck in local maximas where little or no improvements are made for ridiculously long periods of time. It's like getting to the top of the hill. There might be a mountain in the distance, but there is no path that leads to it without first going down off the current hill.