Skip existing tests that have consistently been failing in Rolling-CI

[d3r3kk]

Many tests are still failing in our rolling CI builds that have been failing for longer than our ADO pipeline has been running them.

To facilitate maintaining current quality levels we must skip these known-to-fail tests until we get time to adequately fix them.

For each known-failing test, add code that will skip for the OS(s), for the python version(s) it is known to fail in only.

Once the Rolling-CI test runs green, ensure all new failures cause the build to fail completely.

[d3r3kk]

Tests disabled to achieve this:

#2544 Returning correct blank lines/unicode chars Win-py2.7
#2545 Autopep 526 Win/macOS-py3.4-py3.5
#2546 ProcessService & unicode under Win-py2.7
#2547 Import from .env file Win-py2.7
#2548 (newer, but still unfixed) Unittests run multiple Win/macOS-py3.4