Yes, that is what I found too. Transparencies have a wider dynamic range and that can cause issues when printing. As for camera scanning and converting, dynamic range of transparencies have not been an issue up to now in my case.
Transparencies don’t need NLP, but they can be improved, at times, following this hint:
Sometimes, and mostly with Kodachromes, the effect of said procedure is less obvious or absent:
Original scan (left) and positive created by NLP from an internegative (right)
Although the lower LH corner seems to be drowned in the screens above,
details can be retrieved, e.g. with local adjustment, as we can see below:
