Enable building of tests
Disable Firestore (to avoid Firestore issues temporarily)
Enable Fake secure storage (prevents errors on certain unit tests dealing with secure storage)
Add "Debug" config (to catch any unit tests issues in both Release and Debug)

When running all our tests, these LevelDB tests were not running and were causing the testing process
to fail, marking these tests as "Not run". Since they are just testing LevelDB, these issues will be
caught downstream in our firebase cpp unit tests anyway.

Death tests that are depending on an assert statement to cause a crash
fail in Release mode because assert has no effect when NDEBUG is defined

We get much more descriptive logs in google tests on using EXPECT_LE
instead of EXPECT_TRUE(x<y). With EXPECT_LE, the logs will print out
the expected and actual value if the test fails. Super helpful for debugging.

On Mac github runners, we noticed timing tests failing because they are running
slightly longer than expected. Nuding the thresholds ever so slightly to account for this.

We noticed race condition (specifically on macs) where persistent auth data
(in Keystore) isn't deleted before the next test picks up in the suite.
If the next test involves a SignIn and if persistent cache already has a valid
user logged in, there is no change in auth or id tokens and our expected listeners
are not triggered. We could possibly run into this on other platforms too, hence
forcefully waiting 200ms after a SignOut for the persistent data to be cleared.

Locale on github linux runners is returning C.UTF-8.
Setting an env variable fixes it.

The upper limit for threshold in previous condition was causing issues on mac
machines on github runners. It wasn't sustainable to keep increasing the threshold.
Using this test condition now which should make sure Sleep is working as expected.

SignOut in a correct manner instead of setting null pointers.
If we do not sign out, there could be information in the persistent cache
that can prevent a subsequent sign in from happening and cause the tests to
run incorrectly (saw errors when running on github mac runners)