I have a solution -- not the most eligant. Workflow setup is to use starts with a Split node. The top Split path contains the entire document.
The lower split paths contain single pages (won't work for a range) designated by a MAX variable. These pages are monochromatized and then saved. One needs to build the number of lower split paths to equal the maximum pages that will ever need to be monochromatized.
The top split path has a Collect node which waits for a minute to allow the lower split paths to complete. Then a series of insert and delete nodes are used. The insert node takes a saved monochromatized page/job and inserts it into the job file after the page number designated by the MAX variable. The delete node then eliminates the extra page (in color) designated by the MAX variable. The number of insert/delete node pairs needs to match the maximum number of pages that will ever be monochromatized.
If the number of pages that need to be monochromatized is less than the maximum number, just repeat the last page number in the csv for remaining entries. The outcome is the last monochromatized page will be inserted multiple times but only appear once in the resultant pdf. See csv example below where page 2 will be processed twice.
