tests: Report more error info in MI/sync/tclk driver API tests

[bkeryan]

What does this Pull Request accomplish?

In initialize_driver_session(), use client::raise_if_error(), which throws an exception that includes the full error message.

Change ASSERT_* to EXPECT_* in some helper functions so that they print more failure output instead of returning. ASSERT_* returns from the current function, so calling it in a helper function doesn't terminate the current test.

Why should this Pull Request be merged?

When initialize_driver_session fails, it doesn't print any useful information about the error:

[ RUN      ] NiFgenDriverApiTest.AllocateNamedWaveform_WriteNamedWaveformF64_WaveformWrittenSuccessfully
/home/runner/work/grpc-device/grpc-device/source/tests/system/nifgen_driver_api_tests.cpp:71: Failure
Value of: status.ok()
  Actual: false
Expected: true
/home/runner/work/grpc-device/grpc-device/source/tests/system/nifgen_driver_api_tests.cpp:101: Failure
Value of: status.ok()
  Actual: false
Expected: true
[  FAILED  ] NiFgenDriverApiTest.AllocateNamedWaveform_WriteNamedWaveformF64_WaveformWrittenSuccessfully (81 ms)

What testing has been done?

Ran nidmm, nidigital, nifgen, niscope, niswitch, and nitclk tests on my test machine. (They passed.)

[bkeryan]

These are not the test failures I wanted to see.

[ RUN      ] NiRFmxLTEDriverApiTests.NBIoTModAccFromExample_FetchData_DataLooksReasonable
D:\a\grpc-device\grpc-device\source\tests\system\nirfmxlte_driver_api_tests.cpp(380): error: Expected: (actualStatus) >= (0), actual: -376649 vs 0
[  FAILED  ] NiRFmxLTEDriverApiTests.NBIoTModAccFromExample_FetchData_DataLooksReasonable (5026 ms)

[ RUN      ] NiRFmxSpecAnRestrictedDriverApiTests.CacheResult_ReturnsNonEmptyResult
unknown file: error: C++ exception with description "-380407: A measurement with the same name is already in progress." thrown in the test body.
[  FAILED  ] NiRFmxSpecAnRestrictedDriverApiTests.CacheResult_ReturnsNonEmptyResult (1272 ms)