transcription import without linesegmentation

For all the comparison tasks, we absolutely need the ability to import a second (third, nth) transcription that has the same line-segmentation that does NOT destroy the previous segmentation and transcription but rather becomes a second transcription linked to the same layout, just like a second automatic transcription is done with a different model. That was all the purpose for the creation of the transcription modal which is otherwise pretty senseless. I am pretty sure this feature existed previously. It should be straightforward through matching lineIDs.