[FEA]: Add ability for user to control nvtx range names through execution policy

### Is this a duplicate?

- [x] I confirmed there appear to be no [duplicate issues](https://github.com/NVIDIA/cccl/issues) for this request and that I agree to the [Code of Conduct](CODE_OF_CONDUCT.md)

### Area

General CCCL

### Is your feature request related to a problem? Please describe.

All (?) CCCL algorithms that launch work are annotated with nvtx ranges. These use by default the unmangled fully-namespaced name (e.g. `cub::CUB_200700_860_NS::detail::for_each::static_kernel user_range` to execution policies and pass that user range into the `_CCCL_NVTX_RANGE_SCOPE` macro.

That macro definition would then need to change to include something like:

```
if (user_range.has_value()) {
__cuda_nvtx3_range.emplace(user_range);
} else {
do the old thing
}
```

### Describe alternatives you've considered

_No response_

### Additional context

_No response_