A lot of miners moved to deepbit when the crashed happen.  That sudden surge made Deepbit even more reliable in terms of very steady payouts (higher % of total network hash = likely to make more blocks per day), so some people haven't gone back to slush's pool.

This kind of situation is part of why I've gone the route of starting another pool.  The odds of passing great pools like slush or Deepbit are stacked against me, but that's not my goal.  Splitting the network into smaller pools is a way to help the network respond to pool crashes. 

With the percentage of miners in Deepbit, if it went down transactions across the network would see a large increase in time to be processed until the miners either switched pools or Deepbit came back online.  Compared to the overall network speed right now, I think Slush's pool is at a good size.  It's impact on the network is reasonable, but not crippling in the event of another crash.