I've tried halftoning with carbon with varied results; high-contrast tissue and a hard dot negative = okay. The soft dot of inkjets can cause dot gain/etch issues which then demands VERY consistent exposure and processing of the carbon tissue.
If you can get negs ripped on an imagesetter with accurate screen angles + frequencies, life will be much easier. It only takes a 1/2 dot misregistration to switch from solid to open centred rosettes.
Several years of prepress and hand-coding postscript taught me this the hard way