Fixes #5313

IMPORTANT! Feedback Please
I didn't know whether it would be desirable, but I think it would be a good idea to show a final table that would merge all available accelerators across all k8s clusters? The user might want to only know the total availability across clusters. I think that might be a helpful summary feature. For instance, for the above image, we would have:

Kubernetes GPUs (all allowed contexts):
GPU: T4, TOTAL_GPUS: 2, TOTAL_FREE_GPUS: 2

EDIT (25.04.27): implemented above feature as suggested in below comment

Some Behavioral Changes:

• specifying an accelerator and the kubernetes context (aka region) at the same time ignored the kubernetes context. Fixed that to take into account the k8s context.
• in kubernetes_catalog. _list_accelerators(), the function was adding accelerator names to the dict and setting its availability to 0. I don't know if this is a real requirement (feedback would be great), but this creates an assertion error when checking the keyset sizes for accelerator counts, capacity, and available dicts when used in conjunction with a quantity filter. Therefore, added in checks to NOT add accelerator names to the dict when capacity is 0.
• added a "total table" feature so that users can see a cumulative sum of GPUs across their registered k8s clusters in allowed contexts.

Tested (run the relevant ones):

• Code formatting: install pre-commit (auto-check on commit) or bash format.sh
• Any manual or new tests for this PR (please specify below)
• All smoke tests: /smoke-test (CI) or pytest tests/test_smoke.py (local)
• Relevant individual tests: /smoke-test -k test_name (CI) or pytest tests/test_smoke.py::test_name (local)
• Backward compatibility: /quicktest-core (CI) or pytest tests/smoke_tests/test_backward_compat.py (local)