I've worked on great codebases that would produce thousands of linter errors (lowball estimation), and on shitty codebases that were 100% compliant.