Here's a question and a theory...
I set my speaker distances manually at first, and with REW now. I've read that people are having Emo-Q Come up with settings of 20-30 feet when the true distance is a fraction of that. What I'm wondering is if by doing that, the UMC-200 is increasing it's resolution.
We know that it only gives settings to within several inches. For example, let's say the UMC-200 only gives feet, and odd decimal tenths, (.1, .3, .5, etc.) but not even decimal tenths. The closest setting is 7.3', which equates to about 7' 4". Well, say your head is actually only 7' 2" away. Combine that error over several speakers, and imaging becomes less than optimal.
What if the actual distances stay the same, but you multiply each distance by 3x in the UMC-200. The delay will still work on all speakers equally, but now it's resolution and accuracy can be increased, because the degree of error will be smaller. I.E. 21.5' would get closer to the actual time alignment we want (about 7' 2" rather than 7' 4").
Does that make sense? Is the UMC-200 that smart? Or is it just bad design and I'm way overthinking it?