packaging: redesign the CI target for Pylint job (8.5, quite a good perf, isn't it? :) )

Let's really consider Pylint as a effective linter during the CI completion: instead of failing (and being allowed to), the Pylint check will succeed when the code rate is better than 7.2/10. (That's fair only, and one could target some improvements in a near future.)

EDIT: Look at this 8.5 objective, now one does not consider imports from qgis modules (that decrease the grade).

There is a brand new .pylintrc file for enhancing the Pylint management.

Related issue: #85 (closed)