Adjust weights for rules according to feedback from survey

[bvobart]

When there are more results from the survey, we can start tweaking the weights on each of the rules, as these are currently all 1.

One thing was already noted in the Reddit discussion: the weight for the rule for how many code quality linters are detected / how many are being used, should be weighted 0. The rule is meant to inform users of the linters that are out there and that we recommend, but should prevent users from adding linters to their project just for the sake of adding linters. We may still score the rules for the analysis that those linters do (e.g. code-quality/pylint/no-issues) at 0% when that linter is not installed in the environment they are running mllint in.