NMEA test had 0/0 SOG/COG and was affected by the patch fixing the TracTrac 0/0 SOG/COG issue;

this led to a follow-up issue in the test readWindVaneNmeaIntoWindTrack because the speed inferred
was less than 0.01kts; I now lowered the threshold below which the bearing confidence goes to 0 to
0.001 (from 0.01), making this test case pass again

java/com.sap.sailing.nmeaconnector.test/build.properties

@@ -2,5 +2,4 @@ source.. = src/,\
            resources/
 output.. = bin/
 bin.includes = META-INF/,\
-               .,\
-               lib/routeconverter.jar
+               .