hey,
you're welcome to provide a solution by either by programming it, by looking for (paying?) a programmer, by starting a kickstarter for doing it.
scribus needs new people getting active! (even if it's not easy to get active in the scribus community...)
many people will be very grateful! really!
now, why scribus can do it for jpg and not for pdf? because it's much easier!
jpg images contain only one image. in the pdf, scribus has to keep most of the items separated (text, lines, images, ...) and mix all that with the pre-calculated rasterization related to the flattening.
it for sure doable, but time consuming... and the programmers who are currently active in scribus are already very busy...