Address timeouts of scheduled CI jobs

[khaeru]

Since about 2026-05-31 many or all of the "pytest" workflow jobs reach the GitHub Actions timeout limit of 6 hours and thus fail. See for instance here.

Prior to that, for instance here:

• The total job runtime was about 4 minutes.
• The longest-duration tests were at most about 3 minutes (157 s setup + 32 s call):

============================= slowest 20 durations =============================
157.74s setup    transport_data/tests/org/test_metadata.py::test_merge_ato
139.94s setup    transport_data/tests/test_ato.py::test_convert0
48.57s call     transport_data/tests/test_cli.py::test_cli[command2]
41.86s call     transport_data/tests/org/test_metadata.py::test_tuewas_cli
33.04s call     transport_data/tests/test_estat.py::test_get[TRAN_HV_MS_FRMOD]
32.75s call     transport_data/tests/org/test_metadata.py::test_merge_ato
26.45s call     transport_data/tests/test_cli.py::test_cli[command11]
24.51s call     transport_data/tests/test_ato.py::test_fetch[False]
23.40s setup    transport_data/tests/test_iso.py::test_generate_all
19.42s call     transport_data/tests/test_oica.py::test_update_registry
13.62s call     transport_data/tests/test_jrc.py::test_convert

To resolve:

• Identify the test or tests that are running long.
• Add timeout markers to force these tests to fail.
• Diagnose and fix the long runtimes.