I would suspect that the polar speed is obtained the same way that any other polar speed is obtained for which there is not a discrete TWS value in the polar table (for example, at 7 kts TWS). The VPP that predicated the polar table has the information although it may not be visible in the curves depicted in the polar diagram.

Try this to reveal the polar data: bring up your polars in Exp, click "Edit", click "Upwind", position your cursor in the 4 kt TWS position on the bottom axis of the graph, walk the cursor vertically up the graph until you intersect the BSP curve (or the VMG curve or the TWA curve). The 3 items of data in the upper left corner that change as you move the cursor are 1) TWS, 2) TWA and 3) BSP or VMG (depending where you place the cursor). Be careful not to do any clicking or dragging or you will alter your polar; should you mistakenly do this, simply do not save the changes when you exit the "Edit" function.  

Good Luck in the PH-Mac   

Nick
Are you saying that below 6 knots, where that is your lowest TWS in polars, target BSP will be zero, or that Exp extends the curve below 6 knots, so e.g. in TWS 4 knots , target BSP might be 2 knots?

Good comments, but I need more soak and study time to appreciate the finesse here.  I tried first to guesstimate a 4K polar which led me to the views mentioned in this string, which showed me an approach to improve my original guesses to get a more logical array.  One conclusion after all this fiddling is that the present algorithm that does what Nick states is probably as good a guess as any for the purpose at hand.  Light air being what it is!