Here is a thread that leaves only the words that are not capitals.
non capital
It uses the action "Format value" with this format string:
So, what you could do:
Run the Convert>filename-tag
function twice. The first run imports the whole filename into TITLE, the second in ARTIST.
Now replace with an action of the type "Replace" all underscores with a blank.
Then you apply a modified version of the above string that removes all CAPTIAL words from TITLE.
$regexp(%title%,'\b[\u\s]+\b,\s',)
The second action then deals with ARTIST and removes all words that are non-CAPITAL.
$regexp(%title%,'\b[\l\s]+\b,\s',)
This is just an idea, I have not tested it.