[FEA]: Consider extending cuda::launch() config with optional "name"

### 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

CUDA Experimental (cudax)

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

As a user of `cuda::launch(...)` to launch my kernels, I would like to be able to customize the name of the kernel that shows up in the nsys profile because the default name may be indecipherable and unhelpful.

### Describe the solution you'd like

I'd like to extend the configuration object passed to `cuda::launch` to take an optional string-like parameter:

```
auto config = cudax::make_config(hierarchy, cuda::kernel_name{"my custom kernel"});
// assert( config.name() == "my custom kernel");
cudax::launch(stream, config, kernel, ...);
```
In addition to being able to query the name later, I'd like it to be used internally such that when I profile this code with nsys/ncu the kernel's name is `"my custom kernel"`.

### Describe alternatives you've considered

_No response_

### Additional context

I'm not actually sure if this is possible today without new CUPTI/NVTX APIs.

At minimum, we could insert a custom NVTX push/pop range with the custom name, but that isn't the same as literally renaming the kernel itself.

NCU does support kernel renaming, but this is done at runtime via a config file. We'd want a programmatic way of doing the same thing.

https://docs.nvidia.com/nsight-compute/NsightComputeCli/index.html#kernel-renaming