In quite homogeneous projects you have sometimes cases like:
Wouldn't it be nice to be able to have segments such as 2 and 3 filtered out, as they only contain additional text (= translator needs only to delete sth)?
A similar concept would be to have quite similar segments filtered out, not only the repeated (though here CT already does a better job than the competitors). See below (= filter out not only 1 and 2, but also 3) and 4. I assume length difference can be a lever here.
Something like IF x % same words AND length difference > x chars => FILTER to "Similar segments".