[FEA]: Track advanced optimizations for fine-grained segmented sort

### 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

CUB

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

The initial fine-grained segmented-sort contributions should establish the required planning, primitives, and specialized kernels before incorporating optional performance and runtime-orchestration optimizations.

### Describe the solution you'd like

Track the following segmented-sort optimization work:

- [ ] #10945: optimize the runtime-width group merge-sort primitive.
- [ ] #10950: add a fixed-grid tiny-segment sort path.
- [ ] #10951: add graph-capturable runtime orchestration.

This tracking issue can be closed once the sub-issues have been evaluated and the beneficial approaches have been integrated or explicitly ruled out with supporting results.

### Describe alternatives you've considered

### Additional context

_No response_

Start by reading linked issues #10945, #10950, and #10951 to understand the proposed segmented-sort optimizations and their evaluation criteria. Compare the approaches using supporting performance results; this tracking issue is done when beneficial work is integrated or approaches are explicitly ruled out.

Written by the indexing model from the issue text.