Perhaps the most frustrating part about Ruby, to me, 
  is the outrageously outdated state of the current Ruby interpreter.
  <snip>
  I wish we could change all that. I wish that we had a new, 
  exciting Ruby engine that began to rapidly iterate, was extensible in Ruby itself, 
  and really started to live the dream that Smalltalk failed to achieve in the enterprise: 
  being both cutting edge and mainstream. 

I agree with the above sentiment. A good VM is 'the' need for Ruby right now.

This is where i think Rubinius holds great promise. When Engine Yard started supporting it officially, it started looking like we would finally have a good VM 'soon', eliminating the single biggest issue with Ruby at this point. With the recent cut-down in the size of that team, we are kind of at an infection point. Either this project would flicker out (i really hope it doesn't happen) or the community kicks in a big way to help it pick up pace.