Clean up nvtx macros

[PointKernel]

Description

This PR includes several cleanups for the cudf nvtx wrappers:

• Removed the unused NVTX3_FUNC_RANGE macro
• Fixed a typo in the doc
• Added an example in the cudf::thread_range doc
• Updated the NVTX section in the developer guide doc

Checklist

• [x] I am familiar with the Contributing Guidelines.
• [ ] New or existing tests cover these changes.
• [x] The documentation is up to date with these changes.

[PointKernel]

Notes: Users cannot specify the function name with CUDF_FUNC_RANGE: Custom ranges can be done via cudf::thread_range, i.e. hash_join::inner_join:

[davidwendt]

You could always just call nvtxRangePushA(custom-name)' and 'nvtxRangePop()

[jrhemstad]

Also, this reminds me that the vendored version of nvtx in cudf is wildly out of date and should be removed. See https://github.com/rapidsai/cudf/issues/6476

[PointKernel]

/merge