Tests need to be redesigned

As KernelCI keep evolving we have a lot of tests that has their use at beginning, but becoming more burden than useful now.
For example:
https://github.com/kernelci/kernelci-api/pull/512
Is not really required, as now we have model in kernelci-core package, and actually if any test should verify update against the model. Otherwise we have a problem, if someone update model in kernelci-core, he is forced to update artifical lists inside the tests.

Another example, that tests by themself often require more maintenance than functional code on python version upgrade, which is another efforts burner.
This is a proposal, and maybe even rant, and probably starter for discussion.

Start by reviewing the test areas discussed in this issue, the linked kernelci-api pull request #512, and the model in the kernelci-core package. Identify which artificial test lists duplicate the model and which tests create Python-version maintenance overhead. Done requires an agreed redesign scope and corresponding test changes, but this issue does not define that scope.

Written by the indexing model from the issue text.