[CI] Signal test coverage changes (SimpleCov + Undercover gems)

[filipefurtad0]

What we should change and why (this is tech debt)

We should improve usage of simplecov report by signalling devs whether code changes are having a (negative) impact on code coverage. The undercover gem should be able to provide this information, when integrated in the CI, as it needs a build run on new branches so it can be compared with master branch.

Context

This came up before and on last week on delivery circle meeting.

Impact and timeline

Making better use of the simplecov gem data output. A related discussion can be found here.
.