Two things:

1. The author of this article clearly has money on one side of the issue since he's trying to sell a cross-device app development platform.

2. The data he supplies isn't broken down by usage type, so it's a little hard to analyze. What if 50% of the 'app' time is spent playing games? Or 20% of it is spent only in a couple of apps like Facebook? He's suggesting that native apps are supplanting web usage, but the graph actually shows web usage going up. So, are apps complementing or cannibalizing web usage? Hard to tell.