Admittedly I have no idea what I am looking at with the Oracle 4 numbers. I'm sure I have plenty of company in that regard. But in a quick glance it appears the Sumter County Does have very good numbers in terms of isolating their heritage, if not necessarily finding out who they were in quick fashion.
I believe low numbers are what we are looking for, after the @. The lower the number the greater the confidence level that the ancestry can be narrowed to that combination. You'll note that the numbers following the @ are always the highest (worst confidence level) using one population, then drop somewhat when using two populations, then drop some more using three populations, and are lowest when using four populations.
However, some Does only drop to @9 or even @14 level when narrowed all the way to four populations. That means we're still in glorified guesswork territory.
The Sumter County Does, on the other hand, are already at @3 level using one population. Both drop well below @2 using four populations. Sumter Jock Doe is @1.67 when using the four population combo of Danish + Irish + Irish + Spanish Cataluna.
Sumter Jane Doe is even better. Her four population number is all the way down to @1.32 when using the combo of Austrian + Irish + Irish + Irish.
I think @1.00000 etc is the lowest we go here, correct? That would mean absolute confidence level. Maybe not. I'm a math guy but it's other realms.
For reference purposes, of the 14 charts of Oracle 4 that DNA Doe Project has so thoughtfully provided so far among active cases, only one has better numbers than Sumter County Jane Doe. That would be Chattanooga John Doe, who begins below @3 at one population and drops all the way to @1.09 at four populations. There's one more Doe who has similar numbers to Sumter County Jock Doe, in the @1.67 range.
If I got this all wrong, well...it's interesting to look at anyway. There was greater variance than I anticipated.