API tests sometimes fail because we don't trigger a 429

### Context

The failing test looks like:

```
FAIL: test_all_endpoints (data_refinery_api.test.test_api_general.APITestCases)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/home/user/data_refinery_api/test/test_api_general.py", line 339, in test_all_endpoints
self.assertEqual(response.status_code, status.HTTP_429_TOO_MANY_REQUESTS)
AssertionError: 200 != 429
```

### Problem or idea

It only happens every now and then. I suspect we're relying too carefully on timing and it sometimes doesn't work out. It makes sense since the test is about rate limiting requests.

### Solution or next step

Make the test more reliable.

Start in test/test_api_general.py at test_all_endpoints, around line 339, and inspect how the rate-limit request sequence produces the expected 429 response. Make the test reliably assert 429 instead of intermittently receiving 200, then rerun the API tests to confirm the failure no longer depends on timing.

Written by the indexing model from the issue text.