For Kaplan power sample #3, the total reactive power is given in the solutions as sqrt(3)x[P(WB) - P(WA)]. After researching many different sources, including some old circuits II lab material, and the excellent handout recommended by Flyer, I'm still not sure why the wattmeter B power is subtracted from the wattmeter A power. This of course impacts the solution (-Q or leading pf if WB - WA / +Q or lagging pf if WA - WB). Please advise if you have any suggestions.