I am increasingly convinced that the actual implementation of metrics in #subtitling may have a detrimental effect on quality. Looking at mediocre subtitle files with excellent metrics and work with imagination and fluency being punished for technicalities. It's depressing.

And why would such an implementation be bad for quality? If you're a subtitler and you have x time for a task, where would you place your priorities? Delivering a technically excellent file but with mediocre, literal (but not wrong) renderings? Or spend time working on the words?

Yeah, I know, a good subtitler should do all of that in no time. And if the deadline's too tight they should refuse work. Right...