conformance/glsl/misc/large-loop-compile.html should be made more robust

A flake of this test was seen here on Chromium's CI system:
https://ci.chromium.org/p/chromium/builders/luci.chromium.try/win-angle-rel/578
https://chromium-swarm.appspot.com/task?id=42e178fd942aa010&refresh=10&show_raw=1

with the following failure mode:

Traceback (most recent call last):
_RunGpuTest at content\test\gpu\gpu_tests\gpu_integration_test.py:172
self.RunActualGpuTest(url, *args)
RunActualGpuTest at content\test\gpu\gpu_tests\webgl_conformance_integration_test.py:199
getattr(self, test_name)(test_path, *args[1:])
_RunConformanceTest at content\test\gpu\gpu_tests\webgl_conformance_integration_test.py:288
self._CheckTestCompletion()
_CheckTestCompletion at content\test\gpu\gpu_tests\webgl_conformance_integration_test.py:284
self.fail(self._WebGLTestMessages(self.tab))
fail at .swarming_module\bin\Lib\unittest\case.py:410
raise self.failureException(msg)
AssertionError: Shader compilation took an unreasonably long time
FAIL Shader compilation took an unreasonably long time

If it happens that the test machine slows down transiently, it shouldn't cause this test to fail. The test should be made more robust by making multiple measurements.

Start with conformance/glsl/misc/large-loop-compile.html and review how it measures shader compilation time, using the linked Chromium CI failure as context. Make the test use multiple measurements so transient machine slowdowns do not cause failure, then rerun the conformance test and confirm the flaky timeout condition is handled.

Written by the indexing model from the issue text.