My plan was to port an existing application over to RubyMotion so that I could iterate faster on it. After a few days of work, the conclusion I came to was that it wasn't worth porting an existing application over. I didn't try to piece meal it (if someone demonstrates this, it will attract my attention), just a full re-write.
For me the biggest hurdles were debugging. When something goes wrong it has to be relatively easy to debug. I was in on the early days of RubyMotion and found that their debug facilities were not sufficient enough for me (I think I'm more towards the end of the early adopters curve.) It's also my understanding that this has improved drastically since the early days.
With all that said, the number of success stories that are coming out of RubyMotion has my attention (especially 37Signals). I will most certainly consider RubyMotion for any upcoming pet projets. I still won't use it in my business since you need someone versed in Ruby and partially in iOS, so until I have more developers with that skillset it's not worth going with something so esoteric since developers are so hard to come by as is.