The one big benefit of such temporal resolution is that you could use it to calculate vertical growth rate and overlay it onto NEXRAD imagery. This way, you could quite possible tell where rapid thunderstorm (and potentially severe thunderstorm) growth is likely... On the other hand, it would also help show which storms are currently collapsing, so it works either way.
Add that onto the fact that you could see OFB's in real-time, the movement and growth of individual cumulus clouds, etc..
This could also help out areas that have poor 88D coverage - hopefully the spatial resolution of the IR channel will be upped to 1km, and the VIS channel upped even higher... That's probably not going to happen with GOES-13, but it will most likely be included in a future satellite.
Andrew; I think it's eye trickery... I think you're seeing the shadow of the "anvil" on the W/NW side of the surrounding cloud tops.