[FEA]: Evaluate awkward arrays transition to v2

### 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.compute (Python)

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

We don't know how v2 transition affects end users.

### Describe the solution you'd like

This issue can be closed with an evaluation of
- Total time for running Awkward Array tests w/ v1 vs v2 to compare JIT compile time
- Execution time of ADL benchmarks to ensure no performance regressions
- Listing of dependencies and package install size before and after v2

### Describe alternatives you've considered

_No response_

### Additional context

_No response_

Start by locating the Awkward Array tests and ADL benchmarks referenced in the issue, then run the v1 and v2 comparisons. Record JIT compile time, execution time, dependencies, and package install size; done means an evaluation showing whether v2 introduces performance or packaging regressions.

Written by the indexing model from the issue text.