Yes. Probably due to pitching drone ship deck due to heavy seas, the stage reached zero vertical velocity above the deck and had to shut down. Even at the lowest throttle setting the first stage still has a TWR > 1 and cannot hover without accelerating upwards again, so calculating the "suicide burn" timing is really crucial.
No, that's for a single engine. At landing, the first stage is nearly empty, and thus very light. Even the single center engine alone, throttled as low as it will go (40% or something like that, I don't think we have the exact number for the Merlin-D), produces too much thrust to be able to hover. The Grasshopper and F9R testing vehicles carried extra ballast to be able to hover and descend under thrust without cutting their engines.
The landing burn actually used one engine initially, then three, then one again for the last bit. It doesn't burn three all the way through the landing burn.
edit Corrected "TWR > 0" to "TWR > 1".