Test results so far using Dremel (stock) PLA, which seems to conform closely to 1.75mm so I left the Filament Diameter setting alone...
Mult=1.00, Width=0.40 - Resulting walls measure average 0.38mm thick. Surface looks good with an every so slight line texture when scraped with fingernail.
Mult=1.00, Width=0.45 - **Resulting walls measure average 0.45mm thick. Surface looks even smoother with hardly any line texture when scraped with fingernail.
Mult=1.02, Width=0.40 - Resulting walls measure average 0.48mm thick. Surface has noticeable blobs and over oozing pattern.
Mult=1.00, Width=0.41 - Resulting walls measure average 0.42mm thick. Surface very smooth like test #2 with hardly any line texture.
**I'd have to judge #2 (1.00/0.45) as the best surface overall.
So, if I wanted to try to hit 0.40 mm wall thickness on the money I could try either a 0.405 width or 1.005 multiplier but not sure if that many decimal points over to the right would be diminishing returns as far as accuracy goes.
Here is what I find most interesting...most of my full prints have been on the HIGHER end of this spectrum in terms of extrusion width (usually 0.45 to 0.48mm) but I STILL get fairly open gaps between outline shells as well as top surface fill! This leads me to believe there may be something to the claims and complaints about something defective in S3D's slicing algorithms when it comes to these areas. I can even see these gaps in the build preview and predict exactly where they will occur. I need to do some comparisons with the included Dremel slicer to see if there is a difference.
Interesting indeed.