[BUG]: Legacy/per-thread default streams should be singletons

[leofang]

Is this a duplicate?

• [x] I confirmed there appear to be no duplicate issues for this bug and that I agree to the Code of Conduct

Type of Bug

Runtime Error

Component

cuda.core

Describe the bug

When #1483 started I missed that we wanted to expose them as singletons: https://github.com/NVIDIA/cuda-python/blob/264524b3fcf4d32daf983484ee741fca0ae0b7cc/cuda_core/cuda/core/_stream.pyx#L420-L422 whereas the public methods exposed by #1483 would always return new Python instances. We should fix it before v0.6.0 code freeze.

How to Reproduce

n/a

Expected behavior

n/a

Operating System

No response

nvidia-smi output

No response