Unfortunately Prahlad, the way your question was presented was probably not the best - thus the flaming!

It would seem your application is rather specific - most applications would measure the RMS over many cycles of time, so asking our collective experience would probably not gained you anything. Try and reverse engineer the competitor's product to see how they achieve it - they might be looking at the incoming rms power since they know how they're controlling the load and can interpolate the actual applied rms power delivered to the load. Just a thought!

Whilst this chip is fairly expensive, I think it is up to yourself the get one of these devices and to play with it -there's no better way to get experience.