Some things can be solved before compile time, some during compile time, some during startup, and some only by running as long as needed. The earlier the better.
If you like, you can think of compile-time as its own runtime.
I wanted the ability to have control over memory layout and memory allocations, and be tied to a specific model of how structure and design my program determined by the programming language itself, especially ideas which have yet to prove themselves as being "good" (by whatever metric).
I confess I am also not very impressed by the idea of organizing a whole language around ownership. But that doesn't make move semantics a problem.
It is a problem that Rust doesn't let you code your own move. It is another problem that C++ allows it to fail, or to decay to an ordinary copy.