Intermittent failures in mgmt_sso pipeline for random tests due to timing, loading and element visibility

The mgmt_sso CI pipeline currently experiences intermittent test failures across multiple test cases. The failures are non-deterministic — rerunning the same pipeline often passes without code changes. This suggests that the root cause is timing-related, likely due to inconsistent page load times, delayed DOM rendering, or unstable UI states.

=================================== FAILURES =================================== __________________________ test_rancher_sso[firefox] ___________________________

tools/login-test/test_sso_playwright.py:16: in test_rancher_sso
    assert response.status == 200, f"Expected status 200, but got {response.status}"
E   AssertionError: Expected status 200, but got 503
E   assert 503 == 200
E    +  where 503 = <Response url='https://rancher.sylva/' request=<Request url='https://rancher.sylva/' method='GET'>>.status
___________________________ test_vault_sso[firefox] ____________________________
tools/login-test/test_sso_playwright.py:29: in test_vault_sso
    assert response.status == 200, f"Expected status 200, but got {response.status}"
E   AssertionError: Expected status 200, but got 503
E   assert 503 == 200
E    +  where 503 = <Response url='https://vault.sylva/' request=<Request url='https://vault.sylva/' method='GET'>>.status
____________________________ test_flux_sso[firefox] ____________________________
tools/login-test/test_sso_playwright.py:47: in test_flux_sso
    assert response.status == 200, f"Expected status 200, but got {response.status}"
E   AssertionError: Expected status 200, but got 503
E   assert 503 == 200
E    +  where 503 = <Response url='https://flux.sylva/' request=<Request url='https://flux.sylva/' method='GET'>>.status

Recently fixed specifically for neuvector

!5843 (merged)

!5890 (merged)

Similarly like neuvector, it should be required for all test cases under tools/login-test/test_sso_playwright.py