[FEA]: C++29 `<cuda/std/bit>` Implement Bit permutations

### 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.

`` and `cuda/std/simd` currently misses C++29 bit operations https://eel.is/c++draft/bit#permute and https://eel.is/c++draft/simd#bit

### Describe the solution you'd like

Implement:
- [ ] `bit_reverse` (we already have `cuda::bit_reverse`)
- [ ] `bit_repeat`
- [ ] `bit_compress`
- [ ] `bit_expand`
- [ ] `cuda::std::simd` binding

### Describe alternatives you've considered

_No response_

### Additional context

_No response_

Start by reading the existing implementation, including its current bit_reverse support, and then review the cuda/std/simd entry point alongside the linked C++ draft sections. Done means implementing bit_reverse, bit_repeat, bit_compress, and bit_expand, plus the requested cuda::std::simd binding.

Written by the indexing model from the issue text.