Kirk,
#1 - why isn't it possible to code. I would guess that the only thing that is need is to take the hierarchical keywords stored by Lightroom and match them to the PM column view. e.g: a|b|c would be column1 -> column2 -> column3. And if a keyword in that view does not exists, it could be created, automatically
#2 - in prinicple you are right, but (I don't know how it is in other countries), but agencies in Germany don't use the IPTC keyword fields. All information need to be stored inside the caption (terrible but true). Finally, it is much easier to add pre-defined keywords to an image then writing the caption again and again. The disadvantage is, that I have to add the important keywords now to the caption. For example my structure look like:
sport|country|athlet|gender|name
The only real relevant information for the caption would now be the name. I know that I'm able to add single structured keywords to the caption when using the structured keyword panel. But most times I use LR for assiging the keywords, because the tree view and the keyword mechanisms are more compfortable, then the column view (especially on small displays)... sorry ;-)
regards,
Lars