Draft: CI: generic support for sylva-upgrade with multiple starting points

This is a proposal to evolve the CI deployment pipeline definition tooling so that we'll have a single sylva-upgrade scenario that will then generate different job configurations, identifying the initial version depending on the target branch.

At this point it's more to trigger the discussion

In particular, I realized when close to being finished that there is what perhaps is a major drawback: a given job name, such as ☁capo 🚀rke2 🐧ubuntu 🛠oci,ha 🎬sylva-upgrade will not always test the same thing, some of those jobs will be tests from 1.2.1 to release-1.3 and some will be for 1.3.1 to main...