Digital BW, The Print

Having read all posts (thanks) and thought about it, I came to a 
possible workflow based on not using IP and trying to actually 
simulated ABW.

Can someone comments on the below :

All is done in AdobeRVB, G2.2 working space.
Colored parts of the image and BW ones are separated onto 2 
doccuments.

BW doc is converted to a BW profile in the intent of retaining strict 
greyscale values and profiling the output. This is then turned into 
RVB. 
To cope with "neutrality" (whether paper tint, ink tint...) and 
toning capability; a first simple "fill-in with color" layer is added 
with a fusion mask of the RVB channel (that gives a greyscale mask of 
the image to the color layer). First color is thus appplied only to 
highlights and adjustable with opacity variations. This could also 
serve for toning. A second layer of the same nature (fill-in with 
color) is added with a negative of the RVB fusion mask to possibly 
colorize shadows and adjust (if needed) tone variations in this part 
of the image. 

The BW image is now RVB, BW-profile-converted with capabilities for 
toning in all needed ways. That should simulated ABW (and eventually 
IP) to the closest way I could figure out.

RVB doc is simply converted with a usual color profile.

BW doc is paste onto the RVB one (both are RVB tagged), and this can 
be printed "same as source", icm off. Softproofing is no longer 
available, but I can't find a way of preserving it either than full 
color management workflow which will not allow BW profiling from 
which I expect greater linearity and smoothness of the greyscale and 
fine toning capabilities with the layers.

Thinking as much as I could, I'm expecting this would be coping with 
all the profiling issues, neutrality-toning ones, one single pass, 
and contrast balance of the all output. 

Where I'm unsure (because I have limited knwoledge of it) is the 
coping of GCR. My undertsanding of Tyler's post was that with a BW 
profile I'm to expect a more contrasty BW part than RVB part (black 
generation being stronger in the BW part ?). If so the color layers 
would eventually "hide" this.

Please comment/correct if you feel like it.

Olivier