[FEA]: Finer-grain architecture-conditional nv/target support

### 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

libcu++

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

I would like to have a convenient way to guard using arch-conditional features in CUDA kernels.

Today, with `nv/target`, it supports this via `HAS_FEATURE_SM90a`, but I would like finer grained, feature specific queries like `HAS_FEATURE_WGMMA`, etc.

### Describe the solution you'd like

Add feature specific conditionals usable with nv/target.

### Describe alternatives you've considered

_No response_

### Additional context

_No response_